Geospatial Analytics for Social Good

Published

November 29, 2022

Modified

March 6, 2023

1. OVERVIEW

The main aim of this project is to use the water point related data from rural areas at the water point or small water scheme level available from the WPdx Data Repository to explore the applicable geospatial analysis tools in water resources management for Nigeria.

1.1 Problem Statement

To address the issue of providing clean and sustainable water supply to the rural community.

1.2 Objectives

Apply appropriate global and local measures of spatial Association techniques to reveals the spatial patterns of Not Functional water points.

1.3 Scope of Works

Below are the list of tasks to be completed :

  • Import the shapefile into R and save it in a simple feature data frame format with appropriate sf method.

    note : Nigeria (NGA) has 3 Projected Coordinate Systems, EPSG: 26391, 26392, and 26303.

  • Derive the proportion of functional and non-functional water point at LGA level with appropriate tidyr and dplyr methods,

  • Combine the geospatial and aspatial data frame into simple feature data frame.

  • Performe outliers / clusters analysis by using appropriate local measures of spatial association methods.

  • Perform hotspot areas analysis by using appropriate local measures of spatial association methods.

  • Plot 2 main types of maps below :

    1. Thematic Mapping

      Plot maps to show the spatial distribution of functional and non-functional water point rate at LGA level by using appropriate thematic mapping technique provided by tmap package.

    2. Analytical Mapping

      Plot hotspot areas and outliers / clusters maps of functional and non-functional water point rate at LGA level by using appropriate thematic mapping technique provided by tmap package.


2. R PACKAGE REQUIRED

Following are the packages require for this exercise :

  • pacman package : to install and load the following R packages into R environment.

  • sf package :

    • st_read( ) = import and save the shapefiles into simple feature data table.
    • st_geometry( ) = get geometry from an sf object.
    • st_centroid( ) - takes a polygon or multipolygon and returns the geometric center of the bounding box of the polygon or multipolygon as a point.
  • tidyverse package :

    • dplyr:

      • filter( ) - extract water point records of Nigeria.
      • glimpse( ) - list all columns in a data frame.
      • mutate( ) - edit variables’ values.
    • readr :

      • write_rds( ) - save the extracted sf data table into an output file in rds data format.
    • stringr :

      • str_replace( ) - replace some characters with some other characters in a string.
  • tmap package :

    • qtm( ) - to plot quick thematic map.

    • tm_shape( ) - specify the shape object.

  • spdep package :

    • poly2nb( ) - compute contiguity weight matrices for the study area.

    • nb2listw( ) - supplements a neighbours list with spatial weights for the chosen coding scheme.

    • moran.test( ) - for spatial autocorrelation using a spatial weights matrix in weights list form.

    • moran.mc( ) - for permutation test with Moran’s I statistic.

    • geary.test( ) - for spatial autocorrelation using a spatial weights matrix in weights list form.

    • geary.mc( ) - for permutation test with Geary’s C statistic.

    • sp.correlogram( ) - spatial correlograms for Moran’s I and the autocorrelation coefficient.

    • localmoran( ) - to calculate local spatial statistics for each zone based on the spatial weights object used.

    • moran.plot( ) - plot of spatial data against its spatially lagged values.

    • knearneigh ( ) - return matrix with the indices of points belonging to the set of the k nearest neighbours of each other.

    • knn2nb( ) - convert knn object to a neighbours list of class nb.

    • nbdist( ) - return the length of neighbour relationship edges.

    • dnearneigh( ) - derive distance-based weight matrices.

    • localG( ) - calculate local spatial statistic G for each zone based on the spatial weights object used.

  • funModeling package :

    • freq( ) = retrieves the frequency and percentage for input.
  • janitor package :

    • get_dupes( ) = search duplicated records during data wrangling.

2.1 Load R Packages into R Environment

Use the code chunk below.

Show the code
pacman::p_load(sf, tidyverse, tmap, spdep, funModeling, janitor, maps)


3. GEOSPATIAL DATA

3.1 Acquire Data Source

3.2 Import Aspatial Data - Water Point Shapefile

Use the code chunk below.

Note :

  • st_read( ) to import geo_export data set.

  • #| eval: false to display the code chunks without the output.

Show the code
wp <- st_read(dsn = "/jephOstan/ISSS624/class_project/project_1/data/geodata",
              layer = "geo_export",
              crs = 4326) %>%
  filter(clean_coun == "Nigeria")

3.2.1 Review Imported Aspatial Shapefile

-- retrieve geometry list-column

Use the code chunk below.

Note :

st_geometry( ) to get the geometry summary of a class data.frame or sf object.

Show the code
st_geometry(wp)

-- identify associated attribute

Use the code chunk below.

Note :

glimpse( ) to swtich columns to rows, features to columns.

Show the code
glimpse(wp)

Avoid performing transformation 
if going to use st_intersects() 
in geoprocessing stage. This is 
because st_intersects() only 
works correctly if the geospatial 
data are in geographic coordinate 
system (i.e. wgs84)

3.2.2 Write “wp” in RDS Format

Use the code chunk below.

Notes :

  • write_rds( ) to save the extracted sf data table, i.e. “wp” into an output file in rds data format.

The output file is called “wp_nga.rds” and it is saved in geodata sub-folder.”

Show the code
write_rds(wp, "/jephOstan/ISSS624/class_project/project_1/data/geodata/wp_nga.rds")

3.3 Import Geospatial Data - Nigeria LGA Boundary Data

Use the code chunk below.

Notes :

  • st_read( ) to import geo_export data set.
Show the code
nga <- st_read(dsn = "/jephOstan/ISSS624/class_project/project_1/data/geodata",
               layer = "nga_admbnda_adm2_osgof_20190417",
               crs = 4326)

3.3.1 Review Imported Geospatial Shapefile

-- retrieve geometry list-column

Use the code chunk below.

Note :

st_geometry( ) to get the geometry summary of a class data.frame or sf object.

Show the code
st_geometry(nga)

-- identify associated attribute

Use the code chunk below.

Note :

glimpse( ) to swtich columns to rows, features to columns.

Show the code
glimpse(nga)

3.4 Data Wrangling and Exploratory Data Analysis (EDA)

3.4.1 Recode NA Values into String

Use the code chunk below.

Note :

glimpse( ) to recode all the NA values in status_cle field into Unknown.

Show the code
wp_nga <- read_rds("/jephOstan/ISSS624/class_project/project_1/data/geodata/wp_nga.rds") %>%
  mutate(status_cle = replace_na(status_cle, "Unknown"))

3.4.2 Search Duplicate Rows

Use the code chunk below.

Note :

get_dupes( ) to review rows that have duplicates. Specify the data.frame and the variable combination to search for duplicates and get back the duplicated rows.

Assumption :

For the same value of clean_adm2, each observation has unique pair of lat_deg and lon_deg.

Show the code
duplicates <- wp_nga %>% 
                janitor::get_dupes(clean_adm2,lat_deg,lon_deg)

duplicates


3.4.3 Exploratory Data Analysis

Before proceed to analysis stage, sense the data with EDA.

Use the code chunk below.

Note :

freq( ) to display the distribution of status_cle field in wp_nga.

Show the code
freq(data=wp_nga, 
     input = 'status_cle')

Remarks :

During the first round of EDA, noticed there are 2 values of “Non-functional due to dry season”.


Use mutate( ) with str_replace to combine both values, thereafter update the dataset.

Show the code
wp_nga <- wp_nga %>%
  mutate(status_cle = str_replace(status_cle,"Non functional due to dry season"  ,"Non-Functional due to dry season"))

3.4.4 Explore Variables and Values for “Non-Functional”

Use the code chunk below.

Note :

filter( ) to get the data tables based on values associated with “Non-Functional” under “status_clean” variable.

-- retrieve based on “Non-Functional”

Show the code
wp_nfc <- filter(wp_nga, status_cle == "Non-Functional")
wp_nfc

-- retrieve based on “Non-Functional due to dry season”

Show the code
wp_nfcD <- filter(wp_nga, status_cle == "Non-Functional due to dry season")
wp_nfcD

According to the variables’ definition from the waterpointdata.org :

  • usage_cap (usage_capacity) = recommended maximum users per water point :

    • 1,000 people per mechanised well
    • 500 people per hand pump / undefined shallow well (all hand pumps)
    • 400 people per open hand well (rope & bucket)
    • 250 people per tap (tapstand, kiosk, rainwater catchment)
  • local_popu (local_population_1km) = number of people living within a 1km radius of the water point.

  • served_pop (water_point_population) = number of people currently or potentially served by a specific water point.

    • For non-functional water points, this number represents users who would gain access if the point was functional.
  • pressure (pressure_score) = the pressure score ( 0 - 100% ) is calculated based on the ratio of the number of people assigned to that water point over the theoretical maximum population which can be served based on the technology.

    • If a point is serving less than the recommended maximum, the pressure score will be less than 100% (i.e., 250/500 = 0.5).
    • If serving more than the recommended maximum, the pressure score will be over 100% (i.e. 750/500 = 150%).
  • crucialness (crucialness_score) = the crucialness score ( 0 - 100% ) is the ratio is likely current users to the total local population within a 1km radius of the water point.

    • For non-functional water points, the crucialness score shows how important the water point would be if it were to be rehabilitated.

3.4.5 Plot Preliminary Visualisation for Non-Functional Water Points

Show the code
plot(wp_nfc["crucialnes"])

Remarks :

  1. Visually, noticeable of clusters.
    • This suggests new installation or more resources required to upgrade or rehabilitate the existing facility in order to support population for those light.

3.4.6 Plot Preliminary Visualisation for Non-Functional due to Dry Season Water Points

Show the code
plot(wp_nfcD["crucialnes"])

3.4.7 Plot Preliminary Visualisation for Distribution of Urban & Non-Urban

Show the code
plot(wp_nga["is_urban"])


3.5 Water Point Data Extraction

3.5.1 Extract Non-functional Water Point

Use the code chunk below.

Note :

filter( ) to get water points that associated to the variation of “Non-Functional”.

Show the code
wpt_nonfunctional <- wp_nga %>%
  filter(status_cle %in%
           c("Abandoned/Decommissioned", 
             "Abandoned",
             "Non-Functional",
             "Non-Functional due to dry season"))
Show the code
freq(data=wpt_nonfunctional, 
     input = 'status_cle')

3.5.2 Extract Functional Water Point

Use the code chunk below.

Note :

filter( ) to get water points that associated to the variation of “Functional”.

Show the code
wpt_functional <- wp_nga %>%
  filter(status_cle %in%
           c("Functional", 
             "Functional but not in use",
             "Functional but needs repair"))
Show the code
freq(data=wpt_functional, 
     input = 'status_cle')

3.5.3 Extract Unknown Water Point

Use the code chunk below.

Note :

filter( ) to get water points that associated to the variation of “Unknown”.

Show the code
wpt_unknown <- wp_nga %>%
  filter(status_cle == "Unknown")

3.5.4 Perform Point-In-Polygon Count

Use the code chunk below.

Show the code
nga_wp <- nga %>% 
  mutate(`total_wpt` = lengths(
    st_intersects(nga, wp_nga))) %>%
  mutate(`wpt_functional` = lengths(
    st_intersects(nga, wpt_functional))) %>%
  mutate(`wpt_nonFunctional` = lengths(
    st_intersects(nga, wpt_nonfunctional))) %>%
  mutate(`wpt_unknown` = lengths(
    st_intersects(nga, wpt_unknown)))

3.5.5 Add New Variables

Use the code chunk below.

Note :

mutate( ) to add “pct_functional” and “pct_non-functional”.

write_rds( ) to save the sf data table into rds format.

Show the code
nga_wp <- nga_wp %>%
  mutate(pct_functional = `wpt_functional`/`total_wpt`) %>%
  mutate(`pct_non-functional` = `wpt_nonFunctional`/`total_wpt`)
Show the code
write_rds(nga_wp, "/jephOstan/ISSS624/class_project/project_1/data/geodata/nga_wp.rds")

3.5.6 Spatial Distribution of Water Points Visualisation

Show the code
nga_wp <- read_rds("/jephOstan/ISSS624/class_project/project_1/data/geodata/nga_wp.rds")
Show the code
total <- qtm(nga_wp, "total_wpt") + 
  tm_layout(legend.height = 0.3, legend.width = 0.3)

wp_functional <- qtm(nga_wp, "wpt_functional") +
  tm_layout(legend.height = 0.3, legend.width = 0.3)

wp_nonfunctional <- qtm(nga_wp, "wpt_nonFunctional") + 
  tm_layout(legend.height = 0.3, legend.width = 0.3)

unknown <- qtm(nga_wp, "wpt_unknown") + tm_layout(legend.height = 0.3, legend.width = 0.3)

tmap_arrange(total, wp_functional, wp_nonfunctional, unknown, asp=1.5, ncol=2)

3.6 Exploratory Spatial Data Analysis

Determine autocorrelation for non-functional water points data.

3.6.1 Contiguity Weight Matrices (CWM)

Compute Contiguity Weight Matrices (CWM) first before compute the Global Spatial Autocorrelation statistics.

3.6.1.1 compute Contiguity Weight Matrices

Determine the adjacency with 2 methods, Queen and Rook methods and derive Wq and Wr from the methods respectively.

-- compute QUEEN contiguity based neighbours

Use the code chunk below.

Note :

poly2nb( ) to return a list of first order neighbours based on Queen criteria.

Show the code
nga_wpQ <- poly2nb(nga_wp, queen=TRUE)
summary(nga_wpQ)

-- compute ROOK contiguity based neighbours

Use the code chunk below.

Note :

poly2nb( ) to return a list of first order neighbours based on Queen criteria.

Show the code
nga_wpR <- poly2nb(nga_wp, queen = FALSE)
summary(nga_wpR)

-- Compare Queen and Rook Contiguity Matrices

Wq Wr
Number of regions without links 1 link (#86) 1 link (#86)
Number of regions with the most links 14 links (#508) 14 links (#508)
Number of regions with 1 link #138, #560 #138, #560
Average number of links per region 5.736 5.711

Both methods have relatively similar performance, only have 1 region without links.


3.6.1.2 compute Distance-Based Neighbours

Identify neighbours of region points by Euclidean distance with a distance band.

-- compute “coords”

Use the code chunk below.

Note :

cbind( ) to combine both longitude and latitude

Show the code
longitude <- wp_nga$lon_deg
latitude <- wp_nga$lat_deg
coords <- cbind(longitude, latitude)
head(coords)

-- determine the upper limit for distance band

Use the code chunk below.

Note :

knn2nb( ) to convert knn objects that returned by knearneigh( ) into a neighbours list of class nb.

nbdist( ) to return the length of neighbour relationship edges.

unlist( ) to remove the list structure of the returned object.

Show the code
k3 <- knn2nb(knearneigh(coords,k = 3))
k3dists <- unlist(nbdists(k3, coords, longlat = TRUE))
summary(k3dists)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  13.86   34.82   42.98   45.71   56.00   94.18 
Show the code
upper_threshold3 <- unlist(nbdists(k3, coords, longlat = TRUE))

Remarks :

The largest first nearest neighbour distance is 39.085 km, using this as the upper threshold gives certainty that all units will have at least one neighbour.

Show the code
k5 <- knn2nb(knearneigh(coords,k = 5))
k5dists <- unlist(nbdists(k5, coords, longlat = TRUE))
summary(k5dists)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  13.86   39.06   50.48   54.58   65.14  147.41 
Show the code
upper_threshold5 <- unlist(nbdists(k5, coords, longlat = TRUE))
Show the code
k6 <- knn2nb(knearneigh(coords,k = 6))
k6dists <- unlist(nbdists(k6, coords, longlat = TRUE))
summary(k6dists)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  13.86   41.11   52.45   59.26   72.44  192.60 
Show the code
upper_threshold6 <- unlist(nbdists(k6, coords, longlat = TRUE))
Show the code
k8 <- knn2nb(knearneigh(coords,k = 8))
k8dists <- unlist(nbdists(k8, coords, longlat = TRUE))
summary(k8dists)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  13.86   44.66   58.06   69.16   84.49  267.15 
Show the code
upper_threshold8 <- unlist(nbdists(k8, coords, longlat = TRUE))

-- compute fixed distance weight matrix

Use the code chunk below.

Note :

dnearneigh( ) to derive distance-based weight matrices.

Show the code
wm_d39 <- dnearneigh(coords, 0, upper_threshold3, longlat = TRUE)
wm_d39
Show the code
wm_d40 <- dnearneigh(coords, 0, upper_threshold5, longlat = TRUE)
wm_d40
Show the code
wm_d43 <- dnearneigh(coords, 0, upper_threshold6, longlat = TRUE)
wm_d43
Show the code
wm_d45 <- dnearneigh(coords, 0, upper_threshold8, longlat = TRUE)
wm_d45

3.6.1.3 list all fixed distance weight matrix

Show the code
str(wm_d43)

-- plot Fixed Distance Weight Matrix

Show the code
plot(nga_wp$geometry, border="lightgrey")
plot(wm_d43, coords, add = TRUE)
plot(k1, coords, add = TRUE, col = "red", length = 0.08)
Wd=3 Wd=5 Wd=6 Wd=8
Largest first nearest neighbour distance 39.08489 40.08029 42.28107 44.87575

3.6.1.4 compute Adaptive Distance Weight Matrix

Show the code
knn6 <- knn2nb(knearneigh(coords, k=6))
knn6
Neighbour list object:
Number of regions: 49 
Number of nonzero links: 294 
Percentage nonzero weights: 12.2449 
Average number of links: 6 
Non-symmetric neighbours list

-- plot Adaptive Distance Weight Matrix

Show the code
plot(nga_wp$geometry, border="lightgrey")
plot(knn6, coords, pch = 5, cex = 0.3, add = TRUE, col = "red")

3.6.2 Global Spatial Autocorrelation (GSA)

3.6.2.1 compute row-standardised weights matrix

Use the code chunk below.

Note :

nb2listw( ) to assign weights to each neighboring polygon. Each neighboring polygon will be assigned equal weight by explore with different input for style i.e. “W”, “B” and “C”.

Show the code
wm_q <- poly2nb(nga_wp, queen = TRUE)
wm_q
Neighbour list object:
Number of regions: 774 
Number of nonzero links: 4440 
Percentage nonzero weights: 0.7411414 
Average number of links: 5.736434 
1 region with no links:
86
Show the code
set.ZeroPolicyOption(TRUE)
[1] FALSE
Show the code
ngaWp_rswmQW <- nb2listw(wm_q, style = "W", zero.policy = TRUE)
ngaWp_rswmQW
Characteristics of weights list object:
Neighbour list object:
Number of regions: 774 
Number of nonzero links: 4440 
Percentage nonzero weights: 0.7411414 
Average number of links: 5.736434 
1 region with no links:
86

Weights style: W 
Weights constants summary:
    n     nn  S0       S1       S2
W 773 597529 773 285.0658 3198.414
Show the code
set.ZeroPolicyOption(TRUE)
[1] TRUE
Show the code
ngaWp_rswmQB <- nb2listw(wm_q, style = "B", zero.policy = TRUE)
ngaWp_rswmQB
Characteristics of weights list object:
Neighbour list object:
Number of regions: 774 
Number of nonzero links: 4440 
Percentage nonzero weights: 0.7411414 
Average number of links: 5.736434 
1 region with no links:
86

Weights style: B 
Weights constants summary:
    n     nn   S0   S1     S2
B 773 597529 4440 8880 112368
Show the code
set.ZeroPolicyOption(TRUE)
[1] TRUE
Show the code
ngaWp_rswmQC <- nb2listw(wm_q, style = "C", zero.policy = TRUE)
ngaWp_rswmQC
Characteristics of weights list object:
Neighbour list object:
Number of regions: 774 
Number of nonzero links: 4440 
Percentage nonzero weights: 0.7411414 
Average number of links: 5.736434 
1 region with no links:
86

Weights style: C 
Weights constants summary:
    n     nn  S0       S1      S2
C 773 597529 773 269.1572 3405.93

Regardless of which input for the style, all have the same result of 1 region without links while the average number of links is 5.736434.

3.6.2.2 GSA : Moran’s I method

Use the code chunk below.

Note :

moran.test( ) to calculates the Moran’s I Index value and both a a z-score and p-value to evaluate the significance of that Index.

moran.mc( ) to do a permutation test to evaluate the rank of the observed statistic in relation to the statistic of simulated values.

Show the code
moran.test(nga_wp$wpt_nonFunctional, listw = ngaWp_rswmQW, zero.policy = TRUE, na.action = na.omit)

    Moran I test under randomisation

data:  nga_wp$wpt_nonFunctional  
weights: ngaWp_rswmQW  n reduced by no-neighbour observations
  

Moran I statistic standard deviate = 20.043, p-value < 2.2e-16
alternative hypothesis: greater
sample estimates:
Moran I statistic       Expectation          Variance 
      0.433932927      -0.001295337       0.000471516 
Show the code
set.seed(1234)
bperm_m = moran.mc(nga_wp$wpt_nonFunctional,
                 listw = ngaWp_rswmQW,
                 nsim = 999,
                 zero.policy = TRUE,
                 na.action = na.omit)
bperm_m

    Monte-Carlo simulation of Moran I

data:  nga_wp$wpt_nonFunctional 
weights: ngaWp_rswmQW  
number of simulations + 1: 1000 

statistic = 0.43393, observed rank = 1000, p-value = 0.001
alternative hypothesis: greater

Remarks :

  • The value for both actual and random Moran’s I almost the same, i.e. 0.43393, which means the non-functional water points are positively autocorrelated with the data spatially cluster.

  • There is a statistical significant with p-value (2.2e-16) < 0.001, smaller than the alpha value to support the rejection of null hypothesis for the test i.e. the non-functional water points are not randomly spatial.

-- visualise Monte Carlo Moran’s I

Use the code chunk below.

Note :

hist( ) to examine the simulated Moran’s I test statistics by plotting the distribution of the statistical values as a histogram.

Show the code
hist(bperm_m$res, 
     freq=TRUE, 
     breaks=20, 
     xlab="Simulated Moran's I")
abline(v=0, 
       col="red") 

3.6.2.3 GSA : Geary’s C method

Show the code
geary.test(nga_wp$wpt_nonFunctional, listw=ngaWp_rswmQW)

    Geary C test under randomisation

data:  nga_wp$wpt_nonFunctional 
weights: ngaWp_rswmQW 

Geary C statistic standard deviate = 14.457, p-value < 2.2e-16
alternative hypothesis: Expectation greater than statistic
sample estimates:
Geary C statistic       Expectation          Variance 
     0.6170907765      1.0000000000      0.0007014859 
Show the code
set.seed(1234)
bperm_c = geary.mc(nga_wp$wpt_nonFunctional,
               listw = ngaWp_rswmQW, 
               nsim = 999)
bperm_c

    Monte-Carlo simulation of Geary C

data:  nga_wp$wpt_nonFunctional 
weights: ngaWp_rswmQW 
number of simulations + 1: 1000 

statistic = 0.61709, observed rank = 1, p-value = 0.001
alternative hypothesis: greater

Remarks :

  • The value for both actual and random Geary’s C almost the same, i.e. 0.61709, which means the non-functional water points are positively autocorrelated with the data spatially cluster.

  • There is a statistical significant with p-value (2.2e-16) < 0.001, smaller than the alpha value to support the rejection of null hypothesis for the test i.e. the non-functional water points are not randomly spatial.

  • The Geary’s C value is greater than the Moran’s I value. However, since both values are less than 1, the observations tend to be clustered and similar.


3.7 Cluster and Outlier Analysis

Use Local Indicators for Spatial Association (LISA) method, especially local Moran’s I to detect cluster and / or outlier.

3.7.1 Compute Local Moran’s I

Use the code chunk below.

Show the code
fips <- order(nga_wp$shapeName)
localMI <- localmoran(nga_wp$wpt_nonFunctional, ngaWp_rswmQW)
head(localMI)
           Ii          E.Ii       Var.Ii        Z.Ii Pr(z != E(Ii))
1 -0.32365786 -9.995243e-04 1.924638e-01 -0.73547576     0.46204980
2  0.07000542 -4.092463e-05 1.053077e-02  0.68258288     0.49487045
3  1.25819985 -1.627684e-03 4.181728e-01  1.94819847     0.05139122
4 -0.03537489 -5.427505e-05 5.954304e-03 -0.45773361     0.64714384
5  0.01201533 -2.590965e-04 3.988998e-02  0.06145673     0.95099547
6  0.00768085 -1.538445e-07 1.687859e-05  1.86960486     0.06153871

3.7.1.1 list the local Moran matrix derived

Use the code chunk below.

Show the code
printCoefmat(data.frame(localMI[fips,], nga_wp$shapeName[fips]), check.names=FALSE)
             Ii        E.Ii      Var.Ii        Z.Ii Pr.z....E.Ii..
1   -3.2366e-01 -9.9952e-04  1.9246e-01 -7.3548e-01     4.6205e-01
2    7.0005e-02 -4.0925e-05  1.0531e-02  6.8258e-01     4.9487e-01
3    1.2582e+00 -1.6277e-03  4.1817e-01  1.9482e+00     5.1391e-02
4   -3.5375e-02 -5.4275e-05  5.9543e-03 -4.5773e-01     6.4714e-01
5    1.2015e-02 -2.5910e-04  3.9890e-02  6.1457e-02     9.5100e-01
6    7.6808e-03 -1.5384e-07  1.6879e-05  1.8696e+00     6.1539e-02
7    2.3716e-01 -6.6542e-04  8.5226e-02  8.1464e-01     4.1528e-01
8    1.3499e-01 -6.9507e-05  1.3396e-02  1.1669e+00     2.4326e-01
9    5.8469e-01 -3.9167e-04  6.0293e-02  2.3828e+00     1.7183e-02
10   1.9145e-01 -2.2881e-04  2.5098e-02  1.2099e+00     2.2630e-01
11   6.7485e-01 -7.6926e-04  8.4332e-02  2.3265e+00     1.9992e-02
12   1.3484e-01 -9.2780e-04  8.8869e-02  4.5543e-01     6.4880e-01
13   1.6286e-02 -4.0925e-05  6.3021e-03  2.0567e-01     8.3705e-01
14  -3.1655e-02 -2.9456e-05  2.8239e-03 -5.9514e-01     5.5175e-01
15  -3.1637e-01 -1.0254e-02  1.1135e+00 -2.9010e-01     7.7174e-01
16   6.8612e-02 -9.2780e-04  1.4275e-01  1.8406e-01     8.5397e-01
17   2.4270e-02 -1.9868e-05  5.1126e-03  3.3971e-01     7.3408e-01
18   5.8712e-01 -1.1792e-03  1.8138e-01  1.3814e+00     1.6717e-01
19  -8.6368e-03 -8.6620e-05  1.1101e-02 -8.1152e-02     9.3532e-01
20  -2.1125e-02 -1.8249e-05  2.8103e-03 -3.9815e-01     6.9052e-01
21  -6.1300e-02 -1.4475e-04  2.2288e-02 -4.0963e-01     6.8208e-01
22   2.8968e-01 -1.3297e-03  3.4172e-01  4.9782e-01     6.1861e-01
23   8.3857e-01 -1.1920e-03  1.8334e-01  1.9612e+00     4.9856e-02
24   1.5893e-03 -2.7477e-05  5.2960e-03  2.2217e-02     9.8228e-01
25   1.3175e-01 -8.6620e-05  9.5025e-03  1.3525e+00     1.7623e-01
26   8.0063e-01 -9.3914e-04  1.8085e-01  1.8849e+00     5.9446e-02
27   8.4077e-01 -1.6277e-03  2.5025e-01  1.6840e+00     9.2191e-02
28  -1.9940e-03 -1.2099e-02  1.3113e+00  8.8241e-03     9.9296e-01
29   3.0859e-01 -2.2324e-04  2.4486e-02  1.9735e+00     4.8442e-02
30   2.5468e-01 -3.1865e-04  3.0540e-02  1.4592e+00     1.4452e-01
31   9.3035e-03 -6.3380e-06  8.1230e-04  3.2665e-01     7.4393e-01
32   1.0902e+00 -1.5504e-03  3.9834e-01  1.7298e+00     8.3669e-02
33   8.0526e-02 -1.9868e-05  2.5464e-03  1.5962e+00     1.1045e-01
34   2.2617e-01 -2.7477e-05  3.0145e-03  4.1198e+00     3.7925e-05
35  -1.7370e-01 -3.5422e-04  3.8849e-02 -8.7950e-01     3.7913e-01
36   3.6215e+00 -1.7248e-02  1.6251e+00  2.8544e+00     4.3117e-03
37  -9.7994e-02 -6.9507e-05  4.8271e-03 -1.4094e+00     1.5870e-01
38   2.6132e+00 -5.7830e-03  6.3080e-01  3.2975e+00     9.7543e-04
39   1.8731e+00 -1.5753e-02  2.9886e+00  1.0926e+00     2.7457e-01
40   2.9695e+00 -3.9022e-03  7.4920e-01  3.4353e+00     5.9197e-04
41   4.0806e+00 -1.3638e-02  2.0716e+00  2.8446e+00     4.4471e-03
42   7.7463e-01 -8.2075e-03  1.5690e+00  6.2497e-01     5.3199e-01
43   7.5445e-01 -1.3297e-03  3.4172e-01  1.2929e+00     1.9605e-01
44  -9.4511e-02 -2.0041e-04  2.5680e-02 -5.8852e-01     5.5619e-01
45  -4.0892e-01 -1.0618e-03  2.7294e-01 -7.8067e-01     4.3500e-01
46   1.7264e-02 -1.8555e-06  3.5764e-04  9.1300e-01     3.6124e-01
47  -6.0666e-01 -4.4019e-03  6.7490e-01 -7.3309e-01     4.6350e-01
48   3.8136e-02 -1.0903e-05  2.1016e-03  8.3213e-01     4.0533e-01
49  -2.6481e-01 -9.2780e-04  1.0170e-01 -8.2749e-01     4.0796e-01
50  -3.6486e-01 -1.7389e-04  1.6668e-02 -2.8247e+00     4.7319e-03
51   6.4059e-01 -1.0618e-03  1.0169e-01  2.0122e+00     4.4203e-02
52   5.2295e-01 -6.1632e-04  7.8942e-02  1.8634e+00     6.2399e-02
53   5.1817e-01 -5.6910e-04  5.4530e-02  2.2214e+00     2.6324e-02
54   6.8852e-01 -7.1640e-04  7.8541e-02  2.4593e+00     1.3920e-02
55   7.4747e-01 -8.8063e-04  2.2641e-01  1.5727e+00     1.1578e-01
56   5.8669e-01 -1.6277e-03  2.5025e-01  1.1760e+00     2.3958e-01
57   9.1402e-01 -1.2599e-03  2.4254e-01  1.8585e+00     6.3096e-02
58   7.1930e-01 -1.3297e-03  1.4569e-01  1.8880e+00     5.9027e-02
59  -3.0961e-02 -5.2377e-04  8.0617e-02 -1.0720e-01     9.1463e-01
60   3.2899e-03 -3.3373e-07  3.6614e-05  5.4375e-01     5.8662e-01
61   7.1599e-02 -2.9126e-04  4.4841e-02  3.3949e-01     7.3424e-01
62   1.1170e+00 -1.6277e-03  2.0827e-01  2.4512e+00     1.4236e-02
63   5.9556e-01 -6.0715e-04  7.7767e-02  2.1378e+00     3.2531e-02
64  -7.0312e-02 -2.9456e-05  2.8239e-03 -1.3226e+00     1.8597e-01
65   4.0210e-02 -3.1865e-04  4.9056e-02  1.8298e-01     8.5481e-01
66   9.0458e-02 -1.1259e-03  1.4414e-01  2.4123e-01     8.0938e-01
67   6.6358e-02 -1.8555e-06  2.0357e-04  4.6511e+00     3.3019e-06
68   9.9615e-01 -1.1259e-03  4.3469e-01  1.5126e+00     1.3038e-01
69   9.5142e-01 -1.1259e-03  9.5720e-02  3.0788e+00     2.0782e-03
70   1.5763e+00 -2.3901e-03  3.6719e-01  2.6053e+00     9.1806e-03
71   1.3676e+00 -3.5471e-03  3.0082e-01  2.4999e+00     1.2422e-02
72   9.2471e-02 -1.4925e-04  1.4307e-02  7.7435e-01     4.3873e-01
73  -3.7128e-02 -8.3203e-05  1.0663e-02 -3.5875e-01     7.1978e-01
74  -3.1518e-02 -1.2649e-04  1.6210e-02 -2.4656e-01     8.0525e-01
75  -9.2317e-03 -1.9519e-04  2.1411e-02 -6.1757e-02     9.5076e-01
76  -3.8035e-02 -2.3951e-06  2.2963e-04 -2.5098e+00     1.2079e-02
77   7.1490e-01 -1.0618e-03  1.3594e-01  1.9419e+00     5.2155e-02
78   8.4336e-01 -1.6277e-03  1.7829e-01  2.0012e+00     4.5369e-02
79   7.2523e-01 -9.9952e-04  1.5377e-01  1.8520e+00     6.4030e-02
80   5.3888e-01 -4.8031e-04  7.3931e-02  1.9837e+00     4.7294e-02
81   3.4904e+00 -9.4838e-03  1.4466e+00  2.9099e+00     3.6157e-03
82   1.1262e+00 -3.3198e-03  8.5145e-01  1.2241e+00     2.2090e-01
83  -4.1318e-01 -3.9167e-04  7.5464e-02 -1.5026e+00     1.3293e-01
84  -1.6794e-01 -5.4275e-05  8.3578e-03 -1.8364e+00     6.6298e-02
85  -1.0457e-01 -3.8586e-05  4.2332e-03 -1.6066e+00     1.0814e-01
86   0.0000e+00  0.0000e+00  0.0000e+00         NaN            NaN
87   9.7051e-02 -1.4925e-04  1.9126e-02  7.0284e-01     4.8215e-01
88   1.3239e-01 -1.1259e-03  1.4414e-01  3.5167e-01     7.2509e-01
89   8.6819e-01 -5.3489e-03  5.8371e-01  1.1434e+00     2.5289e-01
90   2.8316e-01 -1.6128e-03  1.3704e-01  7.6927e-01     4.4173e-01
91   1.2582e+00 -1.6277e-03  3.1322e-01  2.2510e+00     2.4382e-02
92   1.5674e+00 -3.2090e-03  4.0996e-01  2.4530e+00     1.4167e-02
93   1.5345e-02 -2.7477e-05  3.5215e-03  2.5904e-01     7.9560e-01
94   6.5838e-03 -9.8783e-04  9.4613e-02  2.4616e-02     9.8036e-01
95   7.6830e-02 -6.6449e-05  8.5158e-03  8.3329e-01     4.0468e-01
96   1.0339e-01 -8.3203e-05  9.1276e-03  1.0831e+00     2.7876e-01
97   1.3652e-02 -1.0184e-04  1.5681e-02  1.0983e-01     9.1254e-01
98   1.1441e-01 -8.6620e-05  1.3338e-02  9.9137e-01     3.2151e-01
99   1.7511e+00 -5.0690e-03  7.7665e-01  1.9927e+00     4.6290e-02
100  2.3957e-01 -3.9905e-04  6.1428e-02  9.6821e-01     3.3294e-01
101 -3.3289e-02 -9.9952e-04  1.5377e-01 -8.2343e-02     9.3437e-01
102  1.3164e+00 -1.5358e-03  2.3615e-01  2.7122e+00     6.6840e-03
103  4.0316e-01 -6.6542e-04  6.3753e-02  1.5994e+00     1.0974e-01
104  1.8824e-01 -2.5316e-04  2.4265e-02  1.2100e+00     2.2626e-01
105 -1.4448e-01 -1.0184e-04  1.3051e-02 -1.2638e+00     2.0630e-01
106 -2.7746e-01 -1.2649e-04  4.8883e-02 -1.2544e+00     2.0971e-01
107  2.5014e+00 -3.3198e-03  6.3776e-01  3.1364e+00     1.7103e-03
108 -6.5757e-02 -1.2235e-04  1.5680e-02 -5.2416e-01     6.0017e-01
109 -1.5657e-01 -3.5422e-04  4.5383e-02 -7.3329e-01     4.6338e-01
110 -4.0714e-01 -7.5196e-03  9.5649e-01 -4.0861e-01     6.8283e-01
111 -5.2572e-04 -5.6910e-04  5.4530e-02  1.8579e-04     9.9985e-01
112  1.4369e+00 -5.3489e-03  4.0699e-01  2.2607e+00     2.3777e-02
113 -3.8910e-01 -6.1632e-04  7.8942e-02 -1.3827e+00     1.6677e-01
114  2.8350e-01 -3.9905e-04  5.1123e-02  1.2556e+00     2.0926e-01
115  1.1070e+00 -1.6277e-03  1.5580e-01  2.8086e+00     4.9757e-03
116 -3.9128e-03 -2.7477e-05  4.2313e-03 -5.9730e-02     9.5237e-01
117 -2.2763e-01 -2.2881e-04  5.8867e-02 -9.3725e-01     3.4863e-01
118  2.6518e+00 -7.0234e-03  1.0740e+00  2.5656e+00     1.0298e-02
119  5.0676e-01 -2.2324e-04  2.8605e-02  2.9976e+00     2.7211e-03
120  3.1906e-01 -8.3203e-05  1.6036e-02  2.5203e+00     1.1727e-02
121  6.1173e-01 -1.4014e-03  2.1550e-01  1.3208e+00     1.8658e-01
122  1.0868e+00 -1.6277e-03  2.0827e-01  2.3850e+00     1.7081e-02
123 -1.6043e-01 -2.5910e-04  2.2046e-02 -1.0787e+00     2.8070e-01
124  1.1122e+00 -1.7723e-03  2.2675e-01  2.3395e+00     1.9311e-02
125 -2.0497e-02 -1.4925e-04  1.9126e-02 -1.4713e-01     8.8303e-01
126  2.9959e-01 -6.1632e-04  2.3806e-01  6.1529e-01     5.3837e-01
127  4.5026e-03 -3.3373e-07  5.1394e-05  6.2812e-01     5.2993e-01
128 -3.1431e-02 -2.5910e-04  3.9890e-02 -1.5607e-01     8.7598e-01
129  2.0237e-01 -8.8063e-04  1.1277e-01  6.0526e-01     5.4501e-01
130  3.6342e-02 -1.2649e-04  1.3876e-02  3.0959e-01     7.5687e-01
131  8.8429e-01 -2.5835e-03  3.9683e-01  1.4079e+00     1.5917e-01
132  4.7994e-01 -5.2377e-04  5.0189e-02  2.1447e+00     3.1980e-02
133  5.0409e-01 -7.5901e-04  9.7204e-02  1.6193e+00     1.0539e-01
134  8.0259e-01 -1.4014e-03  2.6973e-01  1.5480e+00     1.2161e-01
135  3.2095e-01 -2.5316e-04  4.8784e-02  1.4543e+00     1.4587e-01
136 -5.6408e-01 -2.2881e-04  5.8867e-02 -2.3240e+00     2.0127e-02
137  1.0900e-01 -3.8586e-05  7.4371e-03  1.2644e+00     2.0609e-01
138  1.4615e-02 -2.3951e-06  1.8538e-03  3.3950e-01     7.3423e-01
139  5.5125e-02 -5.1577e-05  6.6100e-03  6.7866e-01     4.9735e-01
140  9.5921e-01 -1.0618e-03  2.7294e-01  1.8380e+00     6.6055e-02
141 -9.5427e-01 -1.3297e-03  1.2731e-01 -2.6707e+00     7.5682e-03
142 -1.5215e-01 -1.2235e-04  2.3581e-02 -9.8999e-01     3.2218e-01
143  1.0215e+00 -1.4014e-03  2.1550e-01  2.2035e+00     2.7561e-02
144  1.7310e-02 -1.8249e-05  2.3389e-03  3.5830e-01     7.2012e-01
145  5.8823e-01 -8.6966e-04  1.3381e-01  1.6104e+00     1.0730e-01
146  1.2015e+00 -1.6277e-03  1.5580e-01  3.0481e+00     2.3030e-03
147  1.9828e-01 -1.1259e-03  1.2339e-01  5.6767e-01     5.7026e-01
148  6.0353e-01 -5.2080e-03  9.9861e-01  6.0916e-01     5.4242e-01
149  2.5690e-02 -1.8555e-06  3.5764e-04  1.3585e+00     1.7430e-01
150  8.6937e-02 -1.2649e-04  1.3876e-02  7.3910e-01     4.5985e-01
151 -2.5042e-01 -2.9126e-04  3.7319e-02 -1.2948e+00     1.9539e-01
152  5.2452e-03 -1.0903e-05  1.1962e-03  1.5197e-01     8.7921e-01
153  2.4361e-01 -3.6124e-04  1.3957e-01  6.5305e-01     5.1372e-01
154  4.0067e-01 -8.1339e-04  2.0914e-01  8.7790e-01     3.8000e-01
155 -1.3898e-01 -4.3100e-04  5.5216e-02 -5.8963e-01     5.5544e-01
156  5.3574e-01 -1.0618e-03  1.3594e-01  1.4559e+00     1.4541e-01
157  8.9537e-01 -1.4749e-03  1.4120e-01  2.3867e+00     1.6999e-02
158  1.9339e-02 -8.1339e-04  1.5665e-01  5.0915e-02     9.5939e-01
159  7.6650e-01 -1.4014e-03  1.5353e-01  1.9598e+00     5.0021e-02
160  1.2582e+00 -1.6277e-03  2.0827e-01  2.7606e+00     5.7704e-03
161 -8.4890e-03 -1.0903e-05  1.3974e-03 -2.2680e-01     8.2058e-01
162 -6.3606e-02 -5.4275e-05  4.6190e-03 -9.3509e-01     3.4974e-01
163  1.1977e+00 -1.6128e-03  1.5437e-01  3.0525e+00     2.2691e-03
164  2.0702e+00 -5.2080e-03  5.6841e-01  2.7528e+00     5.9085e-03
165  5.5459e-01 -6.1632e-04  9.4853e-02  1.8027e+00     7.1432e-02
166 -2.2400e-01 -1.4749e-03  1.4120e-01 -5.9220e-01     5.5372e-01
167 -2.0141e-01 -2.0041e-04  2.1983e-02 -1.3571e+00     1.7476e-01
168 -6.0268e-02 -6.0715e-04  7.7767e-02 -2.1394e-01     8.3060e-01
169  1.3772e-01 -3.9905e-04  6.1428e-02  5.5729e-01     5.7733e-01
170 -8.3981e-03 -2.3951e-06  3.0697e-04 -4.7919e-01     6.3180e-01
171  3.9793e-01 -1.6916e-03  6.5269e-01  4.9464e-01     6.2085e-01
172  5.9514e-02 -6.6449e-05  1.7098e-02  4.5565e-01     6.4864e-01
173 -8.1927e-02 -6.3380e-06  5.3942e-04 -3.5272e+00     4.1996e-04
174  5.1947e+00 -7.3523e-03  1.1239e+00  4.9069e+00     9.2505e-07
175 -6.0047e-01 -3.2531e-04  6.2682e-02 -2.3971e+00     1.6526e-02
176  4.1374e-01 -3.2531e-04  5.0081e-02  1.8503e+00     6.4277e-02
177 -5.1956e-02 -1.4749e-03  2.2680e-01 -1.0600e-01     9.1558e-01
178 -5.0841e-02 -5.4275e-05  8.3578e-03 -5.5552e-01     5.7854e-01
179  1.4958e-01 -5.1577e-05  4.9446e-03  2.1279e+00     3.3345e-02
180  8.7695e-02 -2.0041e-04  5.1562e-02  3.8708e-01     6.9869e-01
181  2.6769e-01 -3.2531e-04  5.0081e-02  1.1976e+00     2.3106e-01
182  6.3460e-01 -7.8598e-03  8.5553e-01  6.9459e-01     4.8731e-01
183  2.1155e-01 -1.4925e-04  2.8763e-02  1.2482e+00     2.1195e-01
184  2.1926e-02 -1.0562e-04  2.0355e-02  1.5442e-01     8.7728e-01
185  1.3996e+00 -5.9315e-03  7.5570e-01  1.6168e+00     1.0592e-01
186 -1.0664e+00 -1.3297e-03  2.0450e-01 -2.3552e+00     1.8515e-02
187  6.1959e-01 -1.2468e-03  1.3661e-01  1.6797e+00     9.3016e-02
188  6.6563e-01 -1.5358e-03  1.4701e-01  1.7400e+00     8.1854e-02
189  8.4749e-01 -1.4014e-03  2.1550e-01  1.8286e+00     6.7455e-02
190  9.4563e-01 -1.6277e-03  2.0827e-01  2.0757e+00     3.7926e-02
191  6.9021e-01 -1.4749e-03  1.4120e-01  1.8408e+00     6.5657e-02
192 -7.7908e-01 -5.2377e-04  8.0617e-02 -2.7421e+00     6.1053e-03
193  7.7086e-01 -1.3297e-03  2.0450e-01  1.7076e+00     8.7716e-02
194  9.2093e-01 -1.5504e-03  2.3838e-01  1.8894e+00     5.8840e-02
195  7.6545e-01 -1.5504e-03  3.9834e-01  1.2153e+00     2.2427e-01
196 -3.6995e-01 -1.7723e-03  1.6962e-01 -8.9396e-01     3.7134e-01
197  3.0774e-01 -4.3874e-04  8.4529e-02  1.0600e+00     2.8916e-01
198  2.7947e-01 -4.3874e-04  6.7535e-02  1.0771e+00     2.8144e-01
199  3.0774e-01 -4.3874e-04  3.7324e-02  1.5952e+00     1.1068e-01
200  2.5330e-01 -5.2377e-04  6.7093e-02  9.7994e-01     3.2712e-01
201 -1.1965e-01 -8.1339e-04  1.2516e-01 -3.3591e-01     7.3694e-01
202 -6.6417e-02 -2.2324e-04  2.8605e-02 -3.9138e-01     6.9552e-01
203 -9.9313e-03 -6.3380e-06  6.9535e-04 -3.7638e-01     7.0663e-01
204  3.5951e-01 -9.9952e-04  1.0955e-01  1.0892e+00     2.7606e-01
205  1.7751e-01 -7.1640e-04  1.3799e-01  4.7979e-01     6.3137e-01
206  2.3959e-01 -5.6910e-04  8.7591e-02  8.1148e-01     4.1709e-01
207  3.5680e-01 -1.6277e-03  2.5025e-01  7.1649e-01     4.7369e-01
208  5.9762e-04 -6.9507e-05  8.9077e-03  7.0684e-03     9.9436e-01
209 -5.3116e-02 -2.0041e-04  3.0857e-02 -3.0124e-01     7.6323e-01
210  1.7582e-01 -1.4925e-04  2.2981e-02  1.1608e+00     2.4572e-01
211  1.5230e-01 -3.2531e-04  2.7678e-02  9.1739e-01     3.5894e-01
212  2.4259e-01 -3.6124e-04  3.9618e-02  1.2206e+00     2.2223e-01
213  1.4001e+00 -1.6916e-03  4.3457e-01  2.1265e+00     3.3464e-02
214  5.2532e-01 -7.6926e-04  9.8516e-02  1.6761e+00     9.3712e-02
215  7.5812e-01 -9.3914e-04  1.4449e-01  1.9969e+00     4.5835e-02
216  1.6424e-01 -4.8031e-04  7.3931e-02  6.0582e-01     5.4464e-01
217  5.8749e-01 -8.6966e-04  1.1136e-01  1.7631e+00     7.7886e-02
218  3.6410e-01 -3.1865e-04  4.9056e-02  1.6453e+00     9.9906e-02
219  1.8579e-01 -1.3297e-03  2.0450e-01  4.1378e-01     6.7903e-01
220 -3.2768e-01 -9.3914e-04  1.0294e-01 -1.0184e+00     3.0849e-01
221  2.4600e-02 -1.3162e-03  1.2602e-01  7.3004e-02     9.4180e-01
222  1.7425e-01 -1.3297e-03  1.1302e-01  5.2226e-01     6.0149e-01
223  1.1271e+00 -1.6277e-03  2.0827e-01  2.4733e+00     1.3386e-02
224  9.0870e-01 -4.5316e-03  6.9469e-01  1.0957e+00     2.7322e-01
225  5.8078e-01 -4.8031e-04  4.6027e-02  2.7094e+00     6.7411e-03
226 -3.7568e-01 -5.6910e-04  8.7591e-02 -1.2675e+00     2.0499e-01
227 -1.2497e-01 -1.3297e-03  1.7019e-01 -2.9970e-01     7.6440e-01
228 -1.3336e-01 -3.9905e-04  7.6885e-02 -4.7953e-01     6.3156e-01
229 -2.0255e-01 -6.9507e-05  8.9077e-03 -2.1454e+00     3.1923e-02
230 -7.6106e-01 -1.0452e-02  1.3256e+00 -6.5195e-01     5.1444e-01
231  2.7157e-02 -8.8063e-04  8.4354e-02  9.6537e-02     9.2309e-01
232  5.0818e-01 -1.2599e-03  2.4254e-01  1.0344e+00     3.0093e-01
233  2.6512e-01 -1.9394e-03  1.8558e-01  6.1993e-01     5.3531e-01
234 -5.3344e-01 -8.8063e-04  9.6531e-02 -1.7141e+00     8.6510e-02
235 -1.6336e-02 -1.8555e-06  2.8574e-04 -9.6630e-01     3.3389e-01
236  1.1751e-01 -5.1531e-04  9.9274e-02  3.7459e-01     7.0796e-01
237  1.3117e+00 -3.6636e-03  4.6782e-01  1.9231e+00     5.4466e-02
238 -1.3765e-01 -1.4014e-03  2.6973e-01 -2.6234e-01     7.9306e-01
239  4.2520e-02 -9.8783e-04  1.5197e-01  1.1160e-01     9.1114e-01
240  7.7837e-01 -8.6966e-04  1.3381e-01  2.1302e+00     3.3151e-02
241  1.1322e+00 -1.6277e-03  2.0827e-01  2.4844e+00     1.2978e-02
242 -6.3029e-01 -2.3901e-03  3.6719e-01 -1.0362e+00     3.0011e-01
243  4.9513e-02 -8.6620e-05  1.6694e-02  3.8387e-01     7.0107e-01
244  1.0825e+00 -1.4749e-03  2.2680e-01  2.2762e+00     2.2835e-02
245  3.8218e+00 -8.7432e-03  7.3761e-01  4.4601e+00     8.1903e-06
246  1.0730e+00 -1.5504e-03  2.9837e-01  1.9672e+00     4.9166e-02
247 -1.8102e-01 -6.3380e-06  2.4496e-03 -3.6574e+00     2.5483e-04
248  7.0598e-01 -6.1632e-04  7.8942e-02  2.5149e+00     1.1907e-02
249  1.6376e-01 -2.0041e-04  2.5680e-02  1.0232e+00     3.0623e-01
250  1.2461e+00 -1.6277e-03  2.5025e-01  2.4942e+00     1.2624e-02
251 -5.9560e-02 -3.2531e-04  8.3685e-02 -2.0476e-01     8.3776e-01
252  1.1415e+00 -1.6277e-03  1.7829e-01  2.7074e+00     6.7820e-03
253 -2.9028e-01 -1.5504e-03  1.6983e-01 -7.0062e-01     4.8354e-01
254  2.0483e-01 -3.5422e-04  3.0137e-02  1.1819e+00     2.3724e-01
255  3.8132e-01 -1.0059e-02  2.5625e+00  2.4450e-01     8.0685e-01
256  6.0271e-02 -3.9905e-04  4.3763e-02  2.9001e-01     7.7181e-01
257 -1.8991e-02 -3.2531e-04  3.5679e-02 -9.8820e-02     9.2128e-01
258 -2.6016e-01 -4.8031e-04  5.2671e-02 -1.1315e+00     2.5785e-01
259 -1.5029e-02 -6.9507e-05  6.6634e-03 -1.8326e-01     8.5459e-01
260  2.5243e-01 -8.8063e-04  1.6959e-01  6.1511e-01     5.3848e-01
261  1.2582e+00 -1.6277e-03  2.0827e-01  2.7606e+00     5.7704e-03
262 -1.6611e-02 -8.3203e-05  1.0663e-02 -1.6006e-01     8.7283e-01
263  4.4939e-02 -8.3203e-05  1.0663e-02  4.3600e-01     6.6283e-01
264  1.2976e-01 -3.2531e-04  6.2682e-02  5.1958e-01     6.0336e-01
265  3.3150e-02 -2.3901e-03  4.5958e-01  5.2424e-02     9.5819e-01
266  1.7154e-01 -8.2400e-04  7.8934e-02  6.1351e-01     5.3954e-01
267  4.4473e-01 -9.9952e-04  1.2798e-01  1.2460e+00     2.1277e-01
268  1.9014e-01 -1.2468e-03  1.0598e-01  5.8789e-01     5.5661e-01
269 -3.5537e-02 -5.2377e-04  6.7093e-02 -1.3517e-01     8.9247e-01
270  1.5685e-01 -3.9167e-04  5.0179e-02  7.0197e-01     4.8270e-01
271  1.2582e+00 -1.6277e-03  3.1322e-01  2.2510e+00     2.4382e-02
272  4.8530e-01 -1.6916e-03  4.3457e-01  7.3874e-01     4.6007e-01
273  1.0595e+00 -1.4749e-03  2.2680e-01  2.2278e+00     2.5892e-02
274  1.0249e+00 -1.4749e-03  1.8875e-01  2.3625e+00     1.8151e-02
275 -1.6524e-02 -1.3297e-03  2.0450e-01 -3.3600e-02     9.7320e-01
276 -2.9120e-02 -1.0903e-05  1.3974e-03 -7.7869e-01     4.3616e-01
277  2.3890e-01 -2.5910e-04  3.9890e-02  1.1974e+00     2.3113e-01
278  1.9144e-01 -1.1920e-03  1.8334e-01  4.4987e-01     6.5280e-01
279  2.6967e-01 -2.5910e-04  4.9927e-02  1.2081e+00     2.2703e-01
280  2.2770e-01 -2.0041e-04  2.1983e-02  1.5371e+00     1.2427e-01
281 -2.2712e-01 -7.5901e-04  1.1680e-01 -6.6235e-01     5.0774e-01
282  1.8845e-02 -4.0925e-05  5.2449e-03  2.6077e-01     7.9427e-01
283 -3.8965e-02 -5.6029e-04  8.6235e-02 -1.3078e-01     8.9595e-01
284  2.2390e-03 -1.0903e-05  4.2140e-03  3.4659e-02     9.7235e-01
285  1.2904e-01 -8.2400e-04  1.2679e-01  3.6470e-01     7.1534e-01
286 -9.1998e-03 -1.4925e-04  2.2981e-02 -5.9702e-02     9.5239e-01
287 -4.0001e-01 -2.9126e-04  3.1946e-02 -2.2364e+00     2.5326e-02
288  1.3579e-02 -1.0562e-04  1.3535e-02  1.1762e-01     9.0637e-01
289  6.7715e-02 -1.9868e-05  5.1126e-03  9.4731e-01     3.4348e-01
290  2.5630e-01 -1.6903e-04  1.4384e-02  2.1384e+00     3.2481e-02
291  8.9115e-01 -1.3297e-03  1.2731e-01  2.5013e+00     1.2374e-02
292  8.3080e-01 -1.1920e-03  2.2948e-01  1.7368e+00     8.2423e-02
293  8.3572e-01 -1.4749e-03  1.6158e-01  2.0827e+00     3.7274e-02
294  9.4127e-01 -1.3297e-03  1.7019e-01  2.2849e+00     2.2321e-02
295  1.3369e-01 -1.4925e-04  1.4307e-02  1.1189e+00     2.6316e-01
296 -2.6395e-01 -1.4014e-03  2.1550e-01 -5.6557e-01     5.7169e-01
297 -8.5516e-02 -4.8031e-04  7.3931e-02 -3.1274e-01     7.5448e-01
298  1.2702e+00 -2.3901e-03  4.5958e-01  1.8772e+00     6.0487e-02
299  1.8042e+00 -3.5471e-03  6.8127e-01  2.1902e+00     2.8513e-02
300  5.3311e-01 -3.1865e-04  3.4949e-02  2.8534e+00     4.3255e-03
301  1.1625e+00 -9.2780e-04  1.1880e-01  3.3755e+00     7.3685e-04
302 -2.4336e-01 -4.0925e-05  5.2449e-03 -3.3597e+00     7.8028e-04
303  6.2924e-01 -2.4859e-03  2.7205e-01  1.2112e+00     2.2583e-01
304  7.9307e+00 -5.2576e-02  3.4595e+00  4.2921e+00     1.7698e-05
305  1.2027e+00 -1.1135e-03  2.8623e-01  2.2501e+00     2.4444e-02
306  1.2091e-01 -9.3914e-04  7.9854e-02  4.3119e-01     6.6633e-01
307  4.7157e+00 -2.8615e-02  2.3657e+00  3.0846e+00     2.0384e-03
308  5.1810e-01 -7.1640e-04  6.0928e-02  2.1019e+00     3.5565e-02
309  6.0252e-01 -8.2400e-04  1.2679e-01  1.6944e+00     9.0185e-02
310  7.8575e-01 -1.4014e-03  2.1550e-01  1.6956e+00     8.9958e-02
311  9.7099e-01 -1.4014e-03  2.6973e-01  1.8723e+00     6.1166e-02
312  1.2618e-01 -1.7389e-04  2.2283e-02  8.4644e-01     3.9731e-01
313  2.4753e-01 -1.4925e-04  1.6372e-02  1.9357e+00     5.2905e-02
314  3.1114e-01 -5.6910e-04  8.7591e-02  1.0532e+00     2.9224e-01
315 -9.1066e-01 -1.4749e-03  1.2534e-01 -2.5680e+00     1.0228e-02
316 -2.1566e-02 -5.4275e-05  5.2032e-03 -2.9822e-01     7.6553e-01
317  1.0003e-02 -3.3373e-07  6.4326e-05  1.2473e+00     2.1229e-01
318  7.6232e-02 -6.1632e-04  1.1872e-01  2.2303e-01     8.2351e-01
319 -2.1311e-01 -2.9930e-03  4.5953e-01 -3.0996e-01     7.5659e-01
320  3.4728e-01 -3.9167e-04  3.7536e-02  1.7945e+00     7.2735e-02
321 -8.7294e-03 -3.3373e-07  6.4326e-05 -1.0884e+00     2.7643e-01
322  6.1088e-01 -1.6277e-03  2.5025e-01  1.2244e+00     2.2080e-01
323  3.7480e-01 -8.8063e-04  1.6959e-01  9.1226e-01     3.6163e-01
324 -3.3429e-01 -5.6910e-04  6.2402e-02 -1.3359e+00     1.8157e-01
325  2.8228e-01 -6.6542e-04  1.0240e-01  8.8420e-01     3.7659e-01
326  7.8513e-01 -1.5504e-03  1.4841e-01  2.0421e+00     4.1143e-02
327  2.9551e-01 -1.0618e-03  2.0444e-01  6.5590e-01     5.1189e-01
328  1.1550e-01 -1.4475e-04  2.2288e-02  7.7463e-01     4.3856e-01
329  3.1773e-01 -6.5588e-04  7.1911e-02  1.1873e+00     2.3512e-01
330  1.7514e+00 -1.8549e-03  2.3729e-01  3.5991e+00     3.1928e-04
331  2.8249e-02 -1.9868e-05  2.1798e-03  6.0548e-01     5.4486e-01
332 -7.1532e-02 -4.3100e-04  6.6345e-02 -2.7604e-01     7.8252e-01
333  4.7478e-02 -6.9507e-05  1.0703e-02  4.5959e-01     6.4581e-01
334  6.4821e-03 -8.6620e-05  2.2288e-02  4.3999e-02     9.6491e-01
335  3.1590e-02 -1.9868e-05  2.1798e-03  6.7706e-01     4.9837e-01
336  4.3466e-01 -1.4749e-03  2.8387e-01  8.1859e-01     4.1302e-01
337  1.6600e-01 -2.7477e-05  3.5215e-03  2.7978e+00     5.1457e-03
338 -1.4055e-02 -1.8555e-06  2.3780e-04 -9.1131e-01     3.6213e-01
339  1.0542e+00 -5.6029e-04  6.1436e-02  4.2556e+00     2.0849e-05
340 -4.6502e-02 -6.6542e-04  1.2817e-01 -1.2803e-01     8.9813e-01
341  1.7965e+00 -2.2962e-03  3.5280e-01  3.0285e+00     2.4580e-03
342 -2.0831e-01 -3.2531e-04  6.2682e-02 -8.3073e-01     4.0613e-01
343  4.5466e-01 -1.9519e-04  3.7616e-02  2.3452e+00     1.9016e-02
344  1.0931e-01 -1.0903e-05  2.1016e-03  2.3847e+00     1.7091e-02
345 -9.5441e-02 -1.0497e-03  2.6985e-01 -1.8171e-01     8.5581e-01
346 -3.2929e-01 -4.0925e-05  7.8878e-03 -3.7072e+00     2.0957e-04
347 -1.7755e-01 -1.2162e-05  2.3443e-03 -3.6668e+00     2.4556e-04
348  5.6912e-01 -2.2324e-04  3.4370e-02  3.0710e+00     2.1334e-03
349 -2.7842e-01 -1.6277e-03  2.5025e-01 -5.5330e-01     5.8006e-01
350 -2.2698e-02 -3.9905e-04  3.3949e-02 -1.2102e-01     9.0367e-01
351  2.3604e-01 -7.1640e-04  9.1751e-02  7.8162e-01     4.3444e-01
352 -3.1866e-01 -2.5910e-04  4.9927e-02 -1.4250e+00     1.5417e-01
353  3.4339e-01 -2.2324e-04  4.3019e-02  1.6567e+00     9.7579e-02
354 -2.1322e-02 -1.2162e-05  2.3443e-03 -4.4012e-01     6.5985e-01
355  6.6949e+00 -1.1886e-02  1.5053e+00  5.4665e+00     4.5908e-08
356  5.6083e-03 -1.5384e-07  3.9589e-05  8.9137e-01     3.7273e-01
357 -2.1211e-01 -1.2649e-04  1.2126e-02 -1.9251e+00     5.4218e-02
358  1.6871e-01 -6.6449e-05  1.0232e-02  1.6685e+00     9.5219e-02
359  2.2350e-01 -9.9952e-04  1.2798e-01  6.2755e-01     5.3030e-01
360 -3.6619e-01 -1.5504e-03  1.9839e-01 -8.1866e-01     4.1298e-01
361 -6.0131e-03 -8.2400e-04  7.8934e-02 -1.8470e-02     9.8526e-01
362  4.5740e-01 -1.6916e-03  1.8527e-01  1.0666e+00     2.8617e-01
363 -1.8197e-02 -1.8555e-06  1.5792e-04 -1.4479e+00     1.4764e-01
364  6.1303e-03 -1.0562e-04  1.6263e-02  4.8899e-02     9.6100e-01
365 -2.2750e-02 -5.4275e-05  5.9543e-03 -2.9412e-01     7.6867e-01
366  1.3585e-01 -5.4275e-05  6.9558e-03  1.6296e+00     1.0319e-01
367  1.6027e+00 -2.2324e-04  5.7433e-02  6.6885e+00     2.2541e-11
368  4.3276e-01 -5.6910e-04  8.7591e-02  1.4642e+00     1.4315e-01
369  1.3356e-01 -5.1577e-05  7.9423e-03  1.4992e+00     1.3381e-01
370  1.8036e-01 -2.9126e-04  5.6124e-02  7.6255e-01     4.4573e-01
371  2.2985e-01 -5.2377e-04  8.0617e-02  8.1138e-01     4.1715e-01
372  8.0179e-01 -1.4749e-03  2.8387e-01  1.5076e+00     1.3165e-01
373 -1.3378e-02 -2.3951e-06  2.2963e-04 -8.8267e-01     3.7741e-01
374 -1.9886e-01 -1.1259e-03  2.1678e-01 -4.2469e-01     6.7106e-01
375  2.5139e-02 -1.0903e-05  2.1016e-03  5.4861e-01     5.8327e-01
376  2.9981e-01 -5.6910e-04  7.2897e-02  1.1125e+00     2.6591e-01
377 -2.7264e-01 -1.3875e-03  1.7758e-01 -6.4368e-01     5.1978e-01
378  2.9537e-04 -1.5384e-07  2.3692e-05  6.0714e-02     9.5159e-01
379 -2.3819e-01 -1.4925e-04  2.2981e-02 -1.5702e+00     1.1636e-01
380  1.8019e+00 -2.4859e-03  3.8186e-01  2.9200e+00     3.5001e-03
381  1.0926e+00 -1.4014e-03  3.6011e-01  1.8231e+00     6.8292e-02
382  4.1537e-02 -1.9868e-05  2.5464e-03  8.2353e-01     4.1021e-01
383 -1.3994e-01 -1.3297e-03  1.4569e-01 -3.6315e-01     7.1649e-01
384  2.7125e-01 -2.5316e-04  6.5130e-02  1.0638e+00     2.8740e-01
385 -2.1869e-03 -1.9868e-05  3.8295e-03 -3.5019e-02     9.7206e-01
386  5.1413e-02 -8.1339e-04  1.0416e-01  1.6182e-01     8.7145e-01
387  2.2660e+00 -1.2313e-02  1.5586e+00  1.8249e+00     6.8011e-02
388  1.3567e-01 -1.9868e-05  3.8295e-03  2.1926e+00     2.8333e-02
389 -1.3819e-01 -8.6620e-05  1.3338e-02 -1.1958e+00     2.3176e-01
390  1.5263e-01 -1.2235e-04  1.5680e-02  1.2199e+00     2.2250e-01
391 -1.3166e-01 -2.0041e-04  3.8621e-02 -6.6893e-01     5.0354e-01
392  7.8292e-01 -1.7723e-03  2.7245e-01  1.5033e+00     1.3275e-01
393 -4.1586e-02 -1.2162e-05  1.5588e-03 -1.0530e+00     2.9234e-01
394  3.3748e-01 -1.4014e-03  1.5353e-01  8.6487e-01     3.8711e-01
395  3.6773e-01 -1.4607e-03  1.6002e-01  9.2290e-01     3.5606e-01
396  7.0376e-01 -7.5901e-04  1.9517e-01  1.5947e+00     1.1077e-01
397 -1.3855e+00 -1.1259e-03  2.8941e-01 -2.5733e+00     1.0074e-02
398 -1.3986e-01 -1.5504e-03  1.9839e-01 -3.1053e-01     7.5616e-01
399  1.0373e-01 -2.5910e-04  3.3198e-02  5.7075e-01     5.6817e-01
400  1.2280e+00 -1.6277e-03  2.5025e-01  2.4579e+00     1.3974e-02
401  6.9862e-01 -7.0234e-03  7.6514e-01  8.0671e-01     4.1983e-01
402  1.6859e-01 -1.5358e-03  1.9653e-01  3.8375e-01     7.0117e-01
403  3.2156e-01 -3.2090e-03  4.0996e-01  5.0722e-01     6.1200e-01
404 -8.3787e-02 -1.6903e-04  4.3490e-02 -4.0097e-01     6.8845e-01
405 -9.5898e-01 -9.3914e-04  7.9854e-02 -3.3903e+00     6.9818e-04
406  1.2582e+00 -1.6277e-03  4.1817e-01  1.9482e+00     5.1391e-02
407  2.6952e-01 -8.1339e-04  1.2516e-01  7.6413e-01     4.4479e-01
408  2.4358e+00 -1.4607e-03  2.2462e-01  5.1426e+00     2.7103e-07
409  3.0479e-01 -1.0497e-03  1.1505e-01  9.0170e-01     3.6722e-01
410  1.1181e-01 -2.9126e-04  3.1946e-02  6.2722e-01     5.3052e-01
411  7.3786e-01 -2.1140e-03  3.2487e-01  1.2983e+00     1.9419e-01
412  4.7488e-03 -2.9456e-05  3.2316e-03  8.4055e-02     9.3301e-01
413 -1.8739e-01 -1.2235e-04  1.8840e-02 -1.3644e+00     1.7246e-01
414  2.6079e-01 -7.1640e-04  1.1024e-01  7.8760e-01     4.3093e-01
415 -4.5621e-02 -3.8586e-05  5.9420e-03 -5.9133e-01     5.5430e-01
416  5.2632e-01 -8.6966e-04  6.6469e-02  2.0448e+00     4.0872e-02
417  6.3472e-01 -5.1531e-04  4.3835e-02  3.0341e+00     2.4128e-03
418 -3.6917e-02 -1.2162e-05  1.3343e-03 -1.0103e+00     3.1236e-01
419 -2.1227e-01 -6.1632e-04  9.4853e-02 -6.8723e-01     4.9194e-01
420  2.5970e-01 -3.1865e-04  8.1973e-02  9.0818e-01     3.6378e-01
421  2.7703e-01 -1.2235e-04  1.3422e-02  2.3923e+00     1.6744e-02
422 -2.6231e-02 -1.0562e-04  1.0125e-02 -2.5965e-01     7.9514e-01
423  3.8165e-01 -2.8497e-04  5.4911e-02  1.6299e+00     1.0312e-01
424  2.5215e-01 -4.8031e-04  9.2535e-02  8.3049e-01     4.0626e-01
425  2.2014e-01 -1.4475e-04  2.2288e-02  1.4755e+00     1.4007e-01
426  1.4889e+00 -4.1483e-03  4.5323e-01  2.2178e+00     2.6571e-02
427 -9.4618e-03 -2.3951e-06  4.6165e-04 -4.4025e-01     6.5975e-01
428 -8.3087e-02 -1.4925e-04  1.9126e-02 -5.9971e-01     5.4870e-01
429 -2.3081e-01 -5.6910e-04  2.1983e-01 -4.9106e-01     6.2339e-01
430  4.3336e-01 -4.8031e-04  4.0859e-02  2.1463e+00     3.1853e-02
431  2.9650e-03 -3.3373e-07  6.4326e-05  3.6973e-01     7.1158e-01
432 -8.1591e-02 -1.2162e-05  1.8730e-03 -1.8850e+00     5.9429e-02
433  7.6751e-01 -3.2090e-03  4.0996e-01  1.2037e+00     2.2870e-01
434 -4.4130e-02 -8.3203e-05  9.1276e-03 -4.6103e-01     6.4477e-01
435 -8.1356e-01 -2.3901e-03  3.6719e-01 -1.3387e+00     1.8068e-01
436 -6.4567e-04 -1.5384e-07  2.3692e-05 -1.3262e-01     8.9449e-01
437 -1.7906e-02 -3.1000e-03  3.9608e-01 -2.3526e-02     9.8123e-01
438  2.8901e-02 -5.4386e-06  5.2141e-04  1.2659e+00     2.0554e-01
439  4.1120e-01 -1.5504e-03  3.9834e-01  6.5397e-01     5.1313e-01
440  1.1281e+00 -1.6277e-03  1.2431e-01  3.2043e+00     1.3540e-03
441  1.0935e+00 -1.9394e-03  2.1236e-01  2.3771e+00     1.7448e-02
442 -3.9073e-02 -1.0562e-04  2.7175e-02 -2.3638e-01     8.1314e-01
443  5.7842e-01 -1.1920e-03  2.2948e-01  1.2099e+00     2.2630e-01
444 -3.1227e-01 -4.8031e-04  7.3931e-02 -1.1467e+00     2.5151e-01
445  1.7688e+00 -1.9394e-03  2.4808e-01  3.5552e+00     3.7773e-04
446  3.2529e-01 -9.8783e-04  1.5197e-01  8.3697e-01     4.0261e-01
447  1.2582e+00 -1.6277e-03  4.1817e-01  1.9482e+00     5.1391e-02
448  2.9856e-01 -9.9952e-04  7.6385e-02  1.0839e+00     2.7842e-01
449  6.7138e-02 -7.1640e-04  1.3799e-01  1.8267e-01     8.5506e-01
450 -1.8849e-03 -1.8555e-06  3.5764e-04 -9.9571e-02     9.2068e-01
451  1.2681e-01 -6.5588e-04  1.0094e-01  4.0121e-01     6.8826e-01
452  7.5819e-01 -1.5358e-03  2.3615e-01  1.5634e+00     1.1796e-01
453  1.0092e-01 -1.4014e-03  2.6973e-01  1.9701e-01     8.4382e-01
454  2.6045e-01 -9.2780e-04  2.3853e-01  5.3517e-01     5.9253e-01
455  2.1905e+00 -2.9930e-03  3.8244e-01  3.5469e+00     3.8973e-04
456  3.4467e-01 -1.9519e-04  2.5012e-02  2.1806e+00     2.9213e-02
457 -9.2051e-02 -5.1577e-05  4.9446e-03 -1.3083e+00     1.9076e-01
458  1.0582e+00 -1.5504e-03  2.9837e-01  1.9401e+00     5.2364e-02
459  6.2038e-01 -2.5835e-03  2.4705e-01  1.2534e+00     2.1008e-01
460  1.7886e-03 -8.3203e-05  9.1276e-03  1.9592e-02     9.8437e-01
461  8.6984e-01 -8.8063e-04  1.6959e-01  2.1144e+00     3.4484e-02
462  1.0537e+00 -1.4749e-03  2.2680e-01  2.2157e+00     2.6710e-02
463 -1.5843e-01 -1.6277e-03  2.0827e-01 -3.4359e-01     7.3115e-01
464  1.4196e+00 -2.4859e-03  2.7205e-01  2.7264e+00     6.4020e-03
465 -1.1870e-02 -3.3373e-07  6.4326e-05 -1.4799e+00     1.3890e-01
466  8.4157e-02 -2.0041e-04  2.5680e-02  5.2641e-01     5.9860e-01
467 -2.5992e-01 -1.1792e-03  1.1292e-01 -7.6999e-01     4.4131e-01
468 -1.3204e-01 -5.4275e-05  1.0461e-02 -1.2904e+00     1.9690e-01
469 -2.5568e-01 -9.2780e-04  1.4275e-01 -6.7426e-01     5.0014e-01
470  4.0363e-01 -3.5422e-04  5.4530e-02  1.7300e+00     8.3627e-02
471 -2.3647e-01 -6.1632e-04  4.7118e-02 -1.0865e+00     2.7724e-01
472 -2.0137e-01 -1.1259e-03  1.7320e-01 -4.8117e-01     6.3040e-01
473  1.1675e+00 -1.6277e-03  3.1322e-01  2.0889e+00     3.6716e-02
474  1.8047e-01 -3.2531e-04  3.1178e-02  1.0239e+00     3.0588e-01
475  1.0717e+00 -1.6277e-03  2.0827e-01  2.3518e+00     1.8682e-02
476  1.0032e-02 -5.4386e-06  6.9703e-04  3.8020e-01     7.0380e-01
477  1.1648e+00 -1.4749e-03  1.6158e-01  2.9014e+00     3.7150e-03
478  1.3145e-01 -4.5316e-03  6.9469e-01  1.6314e-01     8.7040e-01
479  6.3142e-01 -1.3297e-03  5.1324e-01  8.8323e-01     3.7711e-01
480  4.2948e-03 -1.5384e-07  2.3692e-05  8.8239e-01     3.7757e-01
481  1.0689e+00 -1.3297e-03  2.5595e-01  2.1153e+00     3.4402e-02
482  2.4269e-01 -1.4475e-04  1.8549e-02  1.7830e+00     7.4589e-02
483 -1.7618e-01 -1.4749e-03  1.6158e-01 -4.3463e-01     6.6383e-01
484 -2.0568e-02 -1.9868e-05  3.8295e-03 -3.3204e-01     7.3986e-01
485 -7.2227e-01 -6.1632e-04  9.4853e-02 -2.3432e+00     1.9121e-02
486  3.5764e-01 -5.6910e-04  8.7591e-02  1.2103e+00     2.2615e-01
487  1.3249e+00 -3.1000e-03  3.9608e-01  2.1101e+00     3.4852e-02
488  8.6679e-02 -2.2324e-04  2.8605e-02  5.1382e-01     6.0738e-01
489  1.2453e-01 -1.4925e-04  1.2701e-02  1.1063e+00     2.6859e-01
490  2.2996e-01 -3.1865e-04  2.7112e-02  1.3986e+00     1.6195e-01
491 -1.1691e+00 -1.4014e-03  2.6973e-01 -2.2484e+00     2.4553e-02
492  1.2582e+00 -1.6277e-03  3.1322e-01  2.2510e+00     2.4382e-02
493  3.2091e-02 -5.1577e-05  4.3894e-03  4.8516e-01     6.2757e-01
494 -2.2500e-02 -2.9456e-05  2.8239e-03 -4.2285e-01     6.7241e-01
495  4.0696e-01 -2.8497e-04  5.4911e-02  1.7379e+00     8.2226e-02
496 -4.1196e-02 -3.8586e-05  7.4371e-03 -4.7726e-01     6.3318e-01
497  2.5996e-01 -1.4014e-03  1.1910e-01  7.5733e-01     4.4885e-01
498  4.1673e-01 -7.6926e-04  6.5421e-02  1.6323e+00     1.0262e-01
499  2.4038e-02 -6.9507e-05  1.3396e-02  2.0828e-01     8.3501e-01
500  1.0360e+00 -1.2599e-03  3.2380e-01  1.8228e+00     6.8330e-02
501  5.3008e-02 -1.5504e-03  2.3838e-01  1.1174e-01     9.1103e-01
502  2.2082e+00 -2.4859e-03  6.3810e-01  2.7674e+00     5.6503e-03
503  5.4171e-02 -1.4749e-03  1.6158e-01  1.3843e-01     8.8990e-01
504  1.8292e-02 -1.2162e-05  1.8730e-03  4.2294e-01     6.7234e-01
505 -1.6085e-01 -1.1135e-03  1.4256e-01 -4.2306e-01     6.7225e-01
506 -1.7727e-01 -8.6620e-05  1.1101e-02 -1.6817e+00     9.2627e-02
507  1.2582e+00 -1.6277e-03  2.5025e-01  2.5184e+00     1.1789e-02
508  4.1411e-01 -1.0497e-03  5.6999e-02  1.7389e+00     8.2048e-02
509  1.2582e+00 -1.6277e-03  2.5025e-01  2.5184e+00     1.1789e-02
510  6.8447e-02 -2.5316e-04  6.5130e-02  2.6920e-01     7.8778e-01
511  8.4196e-01 -6.5588e-04  5.5785e-02  3.5675e+00     3.6035e-04
512  1.1172e+00 -1.5504e-03  2.9837e-01  2.0482e+00     4.0538e-02
513  1.1591e+00 -1.5504e-03  3.9834e-01  1.8389e+00     6.5927e-02
514  2.3732e-01 -3.9905e-04  3.8242e-02  1.2156e+00     2.2413e-01
515  1.0431e+00 -1.2599e-03  1.9378e-01  2.3724e+00     1.7671e-02
516 -1.1309e-03 -1.5384e-07  2.9653e-05 -2.0765e-01     8.3550e-01
517  7.4541e-01 -1.3162e-03  2.5336e-01  1.4835e+00     1.3794e-01
518  9.0273e-02 -9.3914e-04  1.2025e-01  2.6303e-01     7.9253e-01
519 -6.4231e-02 -5.1577e-05  6.6100e-03 -7.8940e-01     4.2988e-01
520  1.9796e-01 -5.6029e-04  5.3686e-02  8.5677e-01     3.9157e-01
521  3.1335e-01 -8.1339e-04  1.5665e-01  7.9377e-01     4.2733e-01
522 -7.1570e-02 -1.8249e-05  1.7496e-03 -1.7106e+00     8.7150e-02
523  1.1464e-01 -1.7389e-04  1.9074e-02  8.3134e-01     4.0578e-01
524  2.8016e-01 -2.0041e-04  3.0857e-02  1.5960e+00     1.1048e-01
525  1.2582e+00 -1.6277e-03  4.1817e-01  1.9482e+00     5.1391e-02
526  1.0897e+00 -1.6277e-03  1.7829e-01  2.5846e+00     9.7504e-03
527  5.1373e-02 -1.9519e-04  2.5012e-02  3.2607e-01     7.4437e-01
528  3.8138e-01 -8.2400e-04  7.8934e-02  1.3604e+00     1.7371e-01
529  1.2072e-02 -6.3380e-06  9.7603e-04  3.8661e-01     6.9905e-01
530 -9.8898e-02 -5.6910e-04  5.4530e-02 -4.2108e-01     6.7370e-01
531  7.6469e-01 -1.1259e-03  1.7320e-01  1.8402e+00     6.5746e-02
532  6.2583e-01 -7.6926e-04  9.8516e-02  1.9964e+00     4.5896e-02
533  9.3026e-02 -1.5504e-03  1.3174e-01  2.6056e-01     7.9443e-01
534  1.5217e-01 -2.9456e-05  4.5360e-03  2.2598e+00     2.3833e-02
535  5.7646e-01 -5.2377e-04  6.7093e-02  2.2275e+00     2.5912e-02
536  8.8679e-01 -9.3914e-04  2.4144e-01  1.8066e+00     7.0818e-02
537  9.2056e-01 -1.4749e-03  1.4120e-01  2.4538e+00     1.4136e-02
538  1.1681e-03 -1.0184e-04  1.9627e-02  9.0651e-03     9.9277e-01
539  4.3614e-02 -2.2881e-04  4.4093e-02  2.0879e-01     8.3461e-01
540 -2.0447e-02 -6.6449e-05  6.3702e-03 -2.5536e-01     7.9845e-01
541  6.7234e-01 -9.3914e-04  1.2025e-01  1.9416e+00     5.2190e-02
542  6.2298e-01 -1.6277e-03  3.1322e-01  1.1160e+00     2.6440e-01
543  4.9715e-01 -5.6910e-04  8.7591e-02  1.6817e+00     9.2621e-02
544  3.2338e-01 -6.6542e-04  6.3753e-02  1.2834e+00     1.9936e-01
545  9.3118e-01 -6.0715e-04  1.5614e-01  2.3581e+00     1.8371e-02
546  1.4691e-02 -3.8586e-05  7.4371e-03  1.7080e-01     8.6438e-01
547  1.6951e-01 -4.7222e-04  9.0976e-02  5.6357e-01     5.7305e-01
548 -2.4197e-01 -4.9318e-03  4.1767e-01 -3.6678e-01     7.1378e-01
549 -6.5236e-02 -5.4386e-06  5.9668e-04 -2.6704e+00     7.5757e-03
550  8.4506e-01 -1.6916e-03  1.4373e-01  2.2335e+00     2.5516e-02
551  9.9888e-03 -1.8249e-05  2.0021e-03  2.2364e-01     8.2303e-01
552  1.9787e-01 -2.5910e-04  3.9890e-02  9.9200e-01     3.2120e-01
553  1.4982e+00 -5.2080e-03  5.6841e-01  1.9941e+00     4.6137e-02
554  1.3832e+00 -2.3901e-03  3.0559e-01  2.5066e+00     1.2191e-02
555  3.6842e-02 -1.9868e-05  2.1798e-03  7.8954e-01     4.2980e-01
556  4.0670e-01 -7.8598e-03  8.5553e-01  4.4820e-01     6.5401e-01
557  2.6021e+00 -5.4918e-03  5.9920e-01  3.3686e+00     7.5543e-04
558  8.4874e-03 -3.6124e-04  3.4620e-02  4.7556e-02     9.6207e-01
559  1.1578e-01 -2.2962e-03  1.5911e-01  2.9601e-01     7.6722e-01
560  6.6227e+00 -3.9022e-03  3.0085e+00  3.8204e+00     1.3321e-04
561  2.7372e-02 -3.2531e-04  3.5679e-02  1.4664e-01     8.8342e-01
562  3.2130e-01 -8.2400e-04  7.0072e-02  1.2169e+00     2.2365e-01
563  1.1029e-01 -4.0925e-05  6.3021e-03  1.3899e+00     1.6457e-01
564  5.2719e-01 -8.2400e-04  6.2983e-02  2.1040e+00     3.5382e-02
565  4.2386e-01 -4.8031e-04  7.3931e-02  1.5606e+00     1.1861e-01
566 -2.0503e-01 -1.7389e-04  4.4739e-02 -9.6850e-01     3.3279e-01
567  2.0166e-01 -2.5316e-04  4.8784e-02  9.1416e-01     3.6063e-01
568  7.3130e-01 -1.1792e-03  1.8138e-01  1.7199e+00     8.5452e-02
569  2.0417e+00 -2.2042e-03  3.3869e-01  3.5121e+00     4.4456e-04
570  7.1880e-03 -2.5910e-04  1.0011e-01  2.3536e-02     9.8122e-01
571  1.2340e+00 -1.6277e-03  2.5025e-01  2.4700e+00     1.3511e-02
572 -1.4659e-01 -1.4925e-04  2.2981e-02 -9.6603e-01     3.3403e-01
573 -2.9739e-01 -3.1865e-04  3.0540e-02 -1.6999e+00     8.9146e-02
574  2.2340e-01 -4.8031e-04  9.2535e-02  7.3597e-01     4.6175e-01
575 -1.7016e-01 -8.3203e-05  9.1276e-03 -1.7802e+00     7.5043e-02
576 -8.4012e-02 -8.1339e-04  1.0416e-01 -2.5779e-01     7.9657e-01
577  5.0585e-01 -1.6916e-03  3.2550e-01  8.8961e-01     3.7368e-01
578  2.9012e-02 -1.2162e-05  2.3443e-03  5.9946e-01     5.4887e-01
579  6.6482e-01 -7.6926e-04  7.3694e-02  2.4518e+00     1.4213e-02
580 -2.8194e-01 -8.6620e-05  1.6694e-02 -2.1814e+00     2.9154e-02
581 -1.6483e-01 -1.2649e-04  1.3876e-02 -1.3982e+00     1.6205e-01
582  3.8743e+00 -6.0818e-03  6.6319e-01  4.7649e+00     1.8891e-06
583 -4.3259e-02 -1.4475e-04  2.2288e-02 -2.8879e-01     7.7274e-01
584 -5.7052e-02 -5.1577e-05  7.9423e-03 -6.3959e-01     5.2244e-01
585  3.6430e-01 -2.9126e-04  2.7916e-02  2.1821e+00     2.9101e-02
586  1.6093e-01 -1.2235e-04  1.8840e-02  1.1733e+00     2.4067e-01
587  3.1158e-02 -3.8586e-05  4.2332e-03  4.7948e-01     6.3160e-01
588  4.8875e-01 -1.1920e-03  1.3062e-01  1.3556e+00     1.7522e-01
589  6.1094e-02 -1.0562e-04  1.3535e-02  5.2604e-01     5.9886e-01
590  1.1394e+00 -1.5504e-03  2.3838e-01  2.3368e+00     1.9449e-02
591 -4.7970e-01 -7.1640e-04  1.3799e-01 -1.2895e+00     1.9724e-01
592  2.9334e-01 -6.6542e-04  7.2956e-02  1.0885e+00     2.7638e-01
593  5.8173e-01 -5.6029e-04  5.3686e-02  2.5131e+00     1.1967e-02
594 -4.9347e-02 -1.0562e-04  1.1586e-02 -4.5747e-01     6.4734e-01
595  2.4317e-01 -5.6910e-04  5.4530e-02  1.0438e+00     2.9659e-01
596 -9.3535e-02 -8.6620e-05  1.1101e-02 -8.8695e-01     3.7511e-01
597  3.8428e-01 -1.1920e-03  2.2948e-01  8.0468e-01     4.2100e-01
598 -1.3621e-01 -2.8497e-04  3.6512e-02 -7.1136e-01     4.7686e-01
599  1.5652e+00 -1.9394e-03  2.9809e-01  2.8704e+00     4.0999e-03
600  1.4039e+00 -3.9022e-03  4.2644e-01  2.1558e+00     3.1097e-02
601  7.3697e-01 -3.1000e-03  2.9629e-01  1.3596e+00     1.7395e-01
602  8.5261e-01 -1.5504e-03  1.6983e-01  2.0727e+00     3.8202e-02
603  9.5775e-01 -1.4749e-03  3.7898e-01  1.5582e+00     1.1920e-01
604  2.2091e-02 -2.9456e-05  3.7751e-03  3.6003e-01     7.1883e-01
605  5.4810e-01 -1.2599e-03  2.4254e-01  1.1155e+00     2.6464e-01
606  7.8103e-02 -1.2649e-04  3.2546e-02  4.3363e-01     6.6456e-01
607  8.6743e-03 -6.3380e-06  9.7603e-04  2.7786e-01     7.8112e-01
608 -3.8818e-01 -6.0715e-04  3.8579e-02 -1.9732e+00     4.8469e-02
609 -2.3074e-02 -1.9868e-05  1.9048e-03 -5.2823e-01     5.9734e-01
610  1.3380e+00 -3.2090e-03  2.4469e-01  2.7115e+00     6.6989e-03
611  6.6415e-01 -7.1640e-04  9.1751e-02  2.1950e+00     2.8166e-02
612  3.8157e-01 -6.0715e-04  1.1696e-01  1.1175e+00     2.6377e-01
613 -2.9816e-01 -1.0618e-03  1.6334e-01 -7.3512e-01     4.6227e-01
614  7.5463e-01 -1.1259e-03  1.7320e-01  1.8160e+00     6.9374e-02
615  5.1528e-01 -9.9952e-04  1.0955e-01  1.5598e+00     1.1880e-01
616  2.1165e-01 -9.9952e-04  2.5695e-01  4.1950e-01     6.7485e-01
617 -2.0130e-02 -5.1577e-05  5.6583e-03 -2.6692e-01     7.8953e-01
618  5.9947e-01 -6.6542e-04  1.0240e-01  1.8754e+00     6.0742e-02
619  6.4364e-01 -8.8063e-04  1.1277e-01  1.9193e+00     5.4942e-02
620  1.3031e+00 -3.7819e-03  4.1335e-01  2.0327e+00     4.2086e-02
621  7.9825e-01 -1.2599e-03  1.9378e-01  1.8163e+00     6.9332e-02
622  6.2773e-01 -8.2400e-04  9.0329e-02  2.0914e+00     3.6495e-02
623  6.9558e-01 -1.6277e-03  2.5025e-01  1.3937e+00     1.6340e-01
624  8.0750e-03 -5.2377e-04  5.0189e-02  3.8383e-02     9.6938e-01
625  4.6860e-01 -5.6029e-04  6.1436e-02  1.8928e+00     5.8383e-02
626  1.1250e-02 -4.0925e-05  3.9234e-03  1.8026e-01     8.5694e-01
627 -2.9406e-02 -6.9507e-05  5.3168e-03 -4.0233e-01     6.8744e-01
628 -1.4238e-01 -8.6620e-05  1.1101e-02 -1.3506e+00     1.7684e-01
629 -7.3679e-03 -1.4925e-04  2.2981e-02 -4.7618e-02     9.6202e-01
630 -5.5478e-03 -5.4275e-05  8.3578e-03 -6.0090e-02     9.5208e-01
631  4.7785e-01 -1.4749e-03  5.6921e-01  6.3532e-01     5.2522e-01
632  2.2707e-01 -1.2649e-04  1.6210e-02  1.7845e+00     7.4339e-02
633  2.1351e-01 -4.3874e-04  5.6206e-02  9.0244e-01     3.6682e-01
634  5.7511e-01 -2.3901e-03  2.6159e-01  1.1291e+00     2.5885e-01
635  1.6371e-01 -8.6966e-04  1.3381e-01  4.4992e-01     6.5277e-01
636  3.1633e-01 -1.7389e-04  2.6774e-02  1.9343e+00     5.3075e-02
637  3.4468e-01 -1.4749e-03  1.4120e-01  9.2121e-01     3.5694e-01
638 -6.3373e-02 -1.2649e-04  1.0764e-02 -6.0960e-01     5.4212e-01
639  1.7885e-01 -8.2400e-04  9.0329e-02  5.9781e-01     5.4997e-01
640  8.3120e+00 -1.8811e-02  2.0249e+00  5.8544e+00     4.7869e-09
641 -2.0520e-03 -3.3373e-07  2.8403e-05 -3.8497e-01     7.0026e-01
642  1.9527e+00 -3.2090e-03  3.5093e-01  3.3016e+00     9.6119e-04
643  5.2612e-01 -9.9952e-04  1.5377e-01  1.3442e+00     1.7888e-01
644  1.1201e+00 -4.7965e-03  6.1179e-01  1.4382e+00     1.5039e-01
645  6.2063e-01 -7.6926e-04  1.4816e-01  1.6144e+00     1.0644e-01
646  7.9197e-01 -1.1920e-03  3.0637e-01  1.4330e+00     1.5186e-01
647  2.0377e+00 -4.2742e-03  5.4546e-01  2.7648e+00     5.6957e-03
648  4.4925e-01 -1.2599e-03  1.3805e-01  1.2125e+00     2.2532e-01
649 -8.4046e-03 -3.3373e-07  5.1394e-05 -1.1723e+00     2.4107e-01
650  2.6986e-02 -3.2531e-04  5.0081e-02  1.2204e-01     9.0287e-01
651  3.6799e-01 -4.8031e-04  7.3931e-02  1.3552e+00     1.7536e-01
652  1.7095e-01 -1.1135e-03  1.2203e-01  4.9256e-01     6.2232e-01
653 -8.3028e-02 -2.9456e-05  4.5360e-03 -1.2324e+00     2.1782e-01
654  6.5987e-01 -1.2599e-03  1.9378e-01  1.5019e+00     1.3313e-01
655  3.8028e-02 -1.2599e-03  1.6127e-01  9.7831e-02     9.2207e-01
656  1.5301e+00 -3.4325e-03  3.7530e-01  2.5032e+00     1.2307e-02
657 -1.1049e-01 -8.1339e-04  8.9166e-02 -3.6728e-01     7.1341e-01
658 -5.1784e-03 -1.2468e-03  1.9176e-01 -8.9783e-03     9.9284e-01
659  9.0495e+00 -1.7248e-02  3.2672e+00  5.0161e+00     5.2740e-07
660 -1.2167e-01 -5.1577e-05  1.3272e-02 -1.0557e+00     2.9113e-01
661 -3.2913e-01 -6.9507e-05  1.3396e-02 -2.8430e+00     4.4688e-03
662 -2.2388e-02 -3.6124e-04  4.6281e-02 -1.0239e-01     9.1845e-01
663 -4.2228e-02 -5.4386e-06  4.6287e-04 -1.9625e+00     4.9702e-02
664  3.1197e-01 -7.0651e-04  1.3608e-01  8.4761e-01     3.9666e-01
665 -1.6296e-03 -2.3951e-06  3.6884e-04 -8.4725e-02     9.3248e-01
666  2.3715e-02 -5.1577e-05  1.3272e-02  2.0631e-01     8.3655e-01
667  1.3349e-02 -1.8555e-06  2.3780e-04  8.6577e-01     3.8661e-01
668  8.4485e-01 -1.0497e-03  1.0054e-01  2.6678e+00     7.6342e-03
669  5.7035e-01 -1.0618e-03  2.0444e-01  1.2638e+00     2.0632e-01
670 -3.1543e-01 -1.7389e-04  6.7196e-02 -1.2162e+00     2.2392e-01
671 -1.1518e-01 -3.9167e-04  5.0179e-02 -5.1242e-01     6.0836e-01
672 -9.6663e-02 -5.6910e-04  8.7591e-02 -3.2469e-01     7.4542e-01
673  8.0602e-02 -9.2780e-04  8.8869e-02  2.7349e-01     7.8448e-01
674  1.3145e-01 -1.6277e-03  1.5580e-01  3.3715e-01     7.3601e-01
675  2.2983e-01 -3.2531e-04  5.0081e-02  1.0284e+00     3.0374e-01
676  9.9270e-01 -1.3297e-03  1.4569e-01  2.6043e+00     9.2072e-03
677  1.4623e+00 -4.2742e-03  4.6692e-01  2.1463e+00     3.1852e-02
678  2.6920e-02 -6.3380e-06  1.6310e-03  6.6675e-01     5.0493e-01
679 -5.8645e-02 -6.6449e-05  1.0232e-02 -5.7910e-01     5.6252e-01
680 -8.5911e-02 -2.5910e-04  2.8419e-02 -5.0808e-01     6.1140e-01
681  1.9560e+00 -4.6631e-03  7.1476e-01  2.3191e+00     2.0389e-02
682  1.0692e+00 -1.4014e-03  2.6973e-01  2.0614e+00     3.9261e-02
683 -6.5842e-02 -6.6542e-04  1.2817e-01 -1.8205e-01     8.5554e-01
684  3.6742e+00 -1.2747e-02  1.2065e+00  3.3566e+00     7.8921e-04
685  2.9738e-01 -1.4014e-03  3.6011e-01  4.9790e-01     6.1855e-01
686  3.0015e-01 -1.1920e-03  2.2948e-01  6.2907e-01     5.2931e-01
687  9.3505e-01 -1.1259e-03  1.2339e-01  2.6651e+00     7.6963e-03
688  2.5622e-01 -2.9126e-04  3.1946e-02  1.4352e+00     1.5124e-01
689  8.9191e-01 -5.1531e-04  6.6011e-02  3.4735e+00     5.1377e-04
690  6.1243e-02 -8.2400e-04  1.5869e-01  1.5580e-01     8.7619e-01
691  1.6924e-01 -8.6966e-04  1.6748e-01  4.1567e-01     6.7765e-01
692  1.4857e-01 -2.7477e-05  4.2313e-03  2.2845e+00     2.2343e-02
693  1.0077e+00 -1.1920e-03  1.5259e-01  2.5827e+00     9.8023e-03
694  1.3377e+00 -2.3901e-03  1.6560e-01  3.2931e+00     9.9086e-04
695 -4.2032e-01 -1.0618e-03  1.6334e-01 -1.0374e+00     2.9957e-01
696 -2.0376e-01 -3.9905e-04  3.0514e-02 -1.1641e+00     2.4436e-01
697  1.0710e+00 -1.4749e-03  2.2680e-01  2.2520e+00     2.4322e-02
698 -1.3067e-01 -2.5835e-03  3.3026e-01 -2.2289e-01     8.2362e-01
699  7.9880e-01 -1.6916e-03  3.2550e-01  1.4031e+00     1.6059e-01
700 -3.1578e-01 -7.0651e-04  1.3608e-01 -8.5411e-01     3.9305e-01
701  2.4912e-02 -5.1577e-05  4.3894e-03  3.7679e-01     7.0633e-01
702 -3.4236e-03 -3.3373e-07  4.2772e-05 -5.2343e-01     6.0067e-01
703  3.3251e-01 -1.1920e-03  2.2948e-01  6.9661e-01     4.8605e-01
704  1.9596e-01 -8.3203e-05  1.6036e-02  1.5482e+00     1.2158e-01
705  8.8341e-01 -1.2599e-03  1.9378e-01  2.0097e+00     4.4462e-02
706 -2.1826e-01 -9.8783e-04  1.2648e-01 -6.1094e-01     5.4124e-01
707  7.3790e-01 -1.5504e-03  2.3838e-01  1.5145e+00     1.2990e-01
708  6.4220e-02 -1.7389e-04  1.3300e-02  5.5837e-01     5.7659e-01
709  6.0549e-02 -1.8249e-05  2.3389e-03  1.2524e+00     2.1044e-01
710 -2.7763e-01 -1.5504e-03  3.9834e-01 -4.3743e-01     6.6180e-01
711 -8.4724e-03 -5.4386e-06  1.0483e-03 -2.6151e-01     7.9370e-01
712 -3.6620e-03 -1.5384e-07  1.6879e-05 -8.9133e-01     3.7275e-01
713 -2.3849e-01 -1.0497e-03  8.9248e-02 -7.9480e-01     4.2673e-01
714 -3.3298e-01 -6.0715e-04  6.6571e-02 -1.2882e+00     1.9767e-01
715  6.7295e-01 -1.5504e-03  2.3838e-01  1.3815e+00     1.6713e-01
716  7.3616e-01 -9.3914e-04  7.9854e-02  2.6084e+00     9.0956e-03
717  7.6129e-01 -1.3297e-03  2.5595e-01  1.5074e+00     1.3171e-01
718 -4.2922e-03 -1.5384e-07  3.9589e-05 -6.8215e-01     4.9515e-01
719  3.2251e-01 -5.2377e-04  4.0046e-02  1.6142e+00     1.0648e-01
720  6.8852e-01 -7.1640e-04  7.8541e-02  2.4593e+00     1.3920e-02
721  8.3329e-02 -1.4475e-04  1.8549e-02  6.1289e-01     5.3995e-01
722  5.3675e-02 -1.7389e-04  2.2283e-02  3.6074e-01     7.1830e-01
723  4.8583e-02 -5.4275e-05  8.3578e-03  5.3201e-01     5.9472e-01
724  7.1148e-01 -9.2780e-04  1.7867e-01  1.6854e+00     9.1907e-02
725  1.1053e-01 -3.5422e-04  3.8849e-02  5.6258e-01     5.7372e-01
726 -3.1866e-01 -6.5588e-04  8.4006e-02 -1.0972e+00     2.7257e-01
727  1.1424e-02 -2.3951e-06  4.6165e-04  5.3181e-01     5.9486e-01
728  5.7671e-01 -7.1640e-04  7.8541e-02  2.0604e+00     3.9362e-02
729  2.7894e-01 -5.6910e-04  6.2402e-02  1.1189e+00     2.6317e-01
730  6.0619e-03 -3.3373e-07  5.1394e-05  8.4562e-01     3.9776e-01
731  3.0813e-01 -7.6926e-04  6.5421e-02  1.2077e+00     2.2716e-01
732  6.7941e-01 -1.4749e-03  1.6158e-01  1.6939e+00     9.0288e-02
733 -1.1686e-02 -2.3951e-06  3.0697e-04 -6.6684e-01     5.0488e-01
734  4.6362e-03 -2.3951e-06  3.6884e-04  2.4153e-01     8.0915e-01
735  9.6682e-01 -2.9930e-03  3.8244e-01  1.5682e+00     1.1683e-01
736  1.0961e+00 -1.2468e-03  3.2043e-01  1.9386e+00     5.2554e-02
737  6.7907e-01 -1.0618e-03  1.6334e-01  1.6828e+00     9.2404e-02
738 -7.6637e-03 -1.0903e-05  8.3406e-04 -2.6499e-01     7.9102e-01
739  7.8882e-01 -1.5504e-03  1.4841e-01  2.0517e+00     4.0203e-02
740  3.4338e-01 -1.2235e-04  1.8840e-02  2.5026e+00     1.2329e-02
741  2.0732e-02 -1.2162e-05  1.0351e-03  6.4477e-01     5.1908e-01
742  4.6167e-02 -5.4386e-06  8.3753e-04  1.5955e+00     1.1061e-01
743 -2.1810e-04 -1.4475e-04  1.5879e-02 -5.8210e-04     9.9954e-01
744  1.3116e-01 -4.0925e-05  7.8878e-03  1.4773e+00     1.3961e-01
745  1.9287e-01 -3.2531e-04  2.7678e-02  1.1612e+00     2.4555e-01
746  3.4420e-01 -2.0041e-04  2.5680e-02  2.1491e+00     3.1624e-02
747  7.0914e-01 -1.1920e-03  1.8334e-01  1.6589e+00     9.7129e-02
748 -7.6487e-02 -5.6029e-04  6.1436e-02 -3.0633e-01     7.5936e-01
749  1.1859e-01 -6.6449e-05  1.0232e-02  1.1731e+00     2.4077e-01
750 -8.0353e-03 -1.9868e-05  2.1798e-03 -1.7168e-01     8.6369e-01
751  8.4291e-01 -2.5835e-03  2.8271e-01  1.5902e+00     1.1180e-01
752  1.2249e-01 -3.6124e-04  6.9603e-02  4.6565e-01     6.4147e-01
753  6.0254e-02 -6.0715e-04  6.6571e-02  2.3588e-01     8.1352e-01
754 -6.1064e-02 -8.3203e-05  1.0663e-02 -5.9055e-01     5.5482e-01
755  5.4543e-03 -1.5384e-07  1.6879e-05  1.3276e+00     1.8430e-01
756  2.2042e+00 -1.2468e-03  1.3661e-01  5.9669e+00     2.4176e-09
757  1.4624e+00 -6.3882e-03  9.7748e-01  1.4856e+00     1.3738e-01
758  1.4254e+00 -4.4019e-03  4.2017e-01  2.2058e+00     2.7401e-02
759  1.7873e+00 -6.7020e-03  7.3036e-01  2.0992e+00     3.5802e-02
760 -1.1477e-01 -2.9456e-05  4.5360e-03 -1.7036e+00     8.8453e-02
761 -3.4257e-01 -7.1640e-04  1.8422e-01 -7.9648e-01     4.2575e-01
762  4.8213e-01 -6.6542e-04  8.5226e-02  1.6538e+00     9.8170e-02
763  1.0412e+00 -1.4014e-03  5.4087e-01  1.4176e+00     1.5631e-01
764  9.9111e-01 -1.1259e-03  1.7320e-01  2.3842e+00     1.7116e-02
765  9.1919e-02 -5.1577e-05  6.6100e-03  1.1312e+00     2.5796e-01
766  1.0843e+00 -1.6277e-03  3.1322e-01  1.9403e+00     5.2346e-02
767  6.9180e-01 -1.2599e-03  1.9378e-01  1.5744e+00     1.1539e-01
768  8.1963e-02 -1.0184e-04  1.1172e-02  7.7642e-01     4.3750e-01
769 -5.1902e-01 -3.9905e-04  7.6885e-02 -1.8704e+00     6.1429e-02
770 -5.5710e-01 -4.8031e-04  6.1529e-02 -2.2440e+00     2.4834e-02
771  1.0395e+01 -1.1260e-02  2.1459e+00  7.1035e+00     1.2159e-12
772 -2.5409e-01 -4.7222e-04  1.8243e-01 -5.9380e-01     5.5265e-01
773 -2.3990e-02 -5.4386e-06  5.2141e-04 -1.0504e+00     2.9354e-01
774 -1.4493e-01 -2.0041e-04  3.8621e-02 -7.3644e-01     4.6146e-01
    nga_wp.shapeName.fips.
1                        1
2                        2
3                        3
4                        4
5                        5
6                        6
7                        7
8                        8
9                        9
10                      10
11                      11
12                      12
13                      13
14                      14
15                      15
16                      16
17                      17
18                      18
19                      19
20                      20
21                      21
22                      22
23                      23
24                      24
25                      25
26                      26
27                      27
28                      28
29                      29
30                      30
31                      31
32                      32
33                      33
34                      34
35                      35
36                      36
37                      37
38                      38
39                      39
40                      40
41                      41
42                      42
43                      43
44                      44
45                      45
46                      46
47                      47
48                      48
49                      49
50                      50
51                      51
52                      52
53                      53
54                      54
55                      55
56                      56
57                      57
58                      58
59                      59
60                      60
61                      61
62                      62
63                      63
64                      64
65                      65
66                      66
67                      67
68                      68
69                      69
70                      70
71                      71
72                      72
73                      73
74                      74
75                      75
76                      76
77                      77
78                      78
79                      79
80                      80
81                      81
82                      82
83                      83
84                      84
85                      85
86                      86
87                      87
88                      88
89                      89
90                      90
91                      91
92                      92
93                      93
94                      94
95                      94
96                      95
97                      96
98                      97
99                      98
100                     99
101                    100
102                    101
103                    102
104                    103
105                    104
106                    105
107                    106
108                    107
109                    108
110                    109
111                    110
112                    111
113                    112
114                    113
115                    114
116                    115
117                    116
118                    117
119                    118
120                    119
121                    120
122                    121
123                    122
124                    123
125                    124
126                    125
127                    126
128                    127
129                    128
130                    129
131                    130
132                    131
133                    132
134                    133
135                    134
136                    135
137                    136
138                    137
139                    138
140                    139
141                    140
142                    141
143                    142
144                    143
145                    144
146                    145
147                    146
148                    147
149                    148
150                    149
151                    150
152                    151
153                    152
154                    153
155                    154
156                    155
157                    156
158                    157
159                    158
160                    159
161                    160
162                    161
163                    162
164                    163
165                    164
166                    165
167                    166
168                    167
169                    168
170                    169
171                    170
172                    171
173                    172
174                    173
175                    174
176                    175
177                    176
178                    177
179                    178
180                    179
181                    180
182                    181
183                    182
184                    183
185                    184
186                    185
187                    186
188                    187
189                    188
190                    189
191                    190
192                    191
193                    192
194                    193
195                    194
196                    195
197                    196
198                    197
199                    198
200                    199
201                    200
202                    201
203                    202
204                    203
205                    204
206                    205
207                    206
208                    207
209                    208
210                    209
211                    210
212                    211
213                    212
214                    213
215                    214
216                    215
217                    216
218                    217
219                    218
220                    219
221                    220
222                    221
223                    222
224                    223
225                    224
226                    225
227                    226
228                    227
229                    228
230                    229
231                    230
232                    231
233                    232
234                    233
235                    234
236                    235
237                    236
238                    237
239                    238
240                    239
241                    240
242                    241
243                    242
244                    243
245                    244
246                    245
247                    246
248                    247
249                    248
250                    249
251                    250
252                    251
253                    252
254                    253
255                    254
256                    255
257                    256
258                    257
259                    258
260                    259
261                    260
262                    261
263                    262
264                    263
265                    264
266                    265
267                    266
268                    267
269                    268
270                    269
271                    270
272                    271
273                    272
274                    273
275                    274
276                    275
277                    276
278                    277
279                    278
280                    279
281                    280
282                    281
283                    282
284                    283
285                    284
286                    285
287                    286
288                    287
289                    288
290                    289
291                    290
292                    291
293                    292
294                    293
295                    294
296                    295
297                    296
298                    297
299                    298
300                    299
301                    300
302                    301
303                    302
304                    303
305                    303
306                    304
307                    305
308                    306
309                    307
310                    308
311                    309
312                    310
313                    311
314                    312
315                    313
316                    314
317                    315
318                    316
319                    317
320                    318
321                    319
322                    320
323                    321
324                    322
325                    323
326                    324
327                    325
328                    326
329                    327
330                    328
331                    329
332                    330
333                    331
334                    332
335                    333
336                    334
337                    335
338                    336
339                    337
340                    338
341                    339
342                    340
343                    341
344                    342
345                    343
346                    344
347                    345
348                    346
349                    347
350                    348
351                    349
352                    350
353                    351
354                    352
355                    353
356                    353
357                    354
358                    355
359                    356
360                    357
361                    358
362                    359
363                    360
364                    361
365                    362
366                    363
367                    364
368                    365
369                    366
370                    367
371                    368
372                    369
373                    370
374                    371
375                    372
376                    373
377                    374
378                    375
379                    376
380                    377
381                    378
382                    379
383                    380
384                    381
385                    382
386                    383
387                    384
388                    385
389                    386
390                    387
391                    388
392                    389
393                    390
394                    391
395                    392
396                    393
397                    394
398                    395
399                    396
400                    397
401                    398
402                    399
403                    400
404                    401
405                    402
406                    403
407                    404
408                    405
409                    406
410                    407
411                    408
412                    409
413                    410
414                    411
415                    412
416                    413
417                    414
418                    415
419                    416
420                    417
421                    418
422                    419
423                    420
424                    421
425                    422
426                    423
427                    424
428                    425
429                    426
430                    427
431                    428
432                    429
433                    430
434                    431
435                    432
436                    433
437                    434
438                    435
439                    436
440                    437
441                    438
442                    439
443                    440
444                    441
445                    442
446                    443
447                    444
448                    445
449                    446
450                    447
451                    448
452                    449
453                    450
454                    451
455                    452
456                    453
457                    454
458                    455
459                    456
460                    457
461                    458
462                    459
463                    460
464                    461
465                    462
466                    463
467                    464
468                    465
469                    466
470                    467
471                    468
472                    469
473                    470
474                    471
475                    472
476                    473
477                    474
478                    475
479                    476
480                    477
481                    478
482                    479
483                    480
484                    481
485                    482
486                    483
487                    484
488                    485
489                    486
490                    487
491                    488
492                    489
493                    490
494                    491
495                    492
496                    493
497                    494
498                    495
499                    496
500                    497
501                    498
502                    499
503                    500
504                    501
505                    502
506                    503
507                    504
508                    505
509                    506
510                    507
511                    508
512                    509
513                    510
514                    511
515                    512
516                    513
517                    514
518                    515
519                    516
520                    516
521                    517
522                    518
523                    519
524                    520
525                    521
526                    522
527                    523
528                    524
529                    525
530                    526
531                    527
532                    528
533                    529
534                    530
535                    531
536                    532
537                    533
538                    534
539                    535
540                    536
541                    537
542                    538
543                    539
544                    540
545                    541
546                    542
547                    542
548                    543
549                    544
550                    545
551                    546
552                    547
553                    548
554                    549
555                    550
556                    551
557                    552
558                    553
559                    554
560                    555
561                    556
562                    557
563                    558
564                    559
565                    560
566                    561
567                    562
568                    563
569                    564
570                    565
571                    566
572                    567
573                    568
574                    569
575                    570
576                    571
577                    572
578                    573
579                    574
580                    575
581                    576
582                    577
583                    578
584                    579
585                    580
586                    581
587                    582
588                    583
589                    584
590                    585
591                    586
592                    587
593                    588
594                    589
595                    590
596                    591
597                    592
598                    593
599                    594
600                    595
601                    596
602                    597
603                    598
604                    599
605                    600
606                    601
607                    602
608                    603
609                    604
610                    605
611                    606
612                    607
613                    608
614                    609
615                    610
616                    611
617                    612
618                    613
619                    614
620                    615
621                    616
622                    617
623                    618
624                    619
625                    620
626                    621
627                    622
628                    623
629                    624
630                    625
631                    626
632                    627
633                    628
634                    629
635                    630
636                    631
637                    632
638                    633
639                    634
640                    635
641                    636
642                    637
643                    638
644                    639
645                    640
646                    641
647                    642
648                    643
649                    644
650                    645
651                    646
652                    647
653                    648
654                    649
655                    650
656                    651
657                    652
658                    653
659                    654
660                    655
661                    656
662                    657
663                    658
664                    659
665                    660
666                    661
667                    662
668                    663
669                    664
670                    665
671                    666
672                    667
673                    668
674                    669
675                    670
676                    671
677                    672
678                    673
679                    674
680                    675
681                    676
682                    677
683                    678
684                    679
685                    680
686                    681
687                    682
688                    683
689                    684
690                    685
691                    686
692                    687
693                    688
694                    688
695                    689
696                    690
697                    691
698                    692
699                    693
700                    694
701                    695
702                    696
703                    697
704                    698
705                    699
706                    700
707                    701
708                    702
709                    703
710                    704
711                    705
712                    706
713                    707
714                    708
715                    709
716                    710
717                    711
718                    712
719                    713
720                    714
721                    715
722                    716
723                    717
724                    718
725                    719
726                    720
727                    721
728                    722
729                    723
730                    724
731                    725
732                    726
733                    727
734                    728
735                    729
736                    730
737                    731
738                    732
739                    733
740                    734
741                    735
742                    736
743                    737
744                    738
745                    739
746                    740
747                    741
748                    742
749                    743
750                    744
751                    745
752                    746
753                    747
754                    748
755                    749
756                    750
757                    751
758                    752
759                    753
760                    754
761                    755
762                    756
763                    757
764                    758
765                    759
766                    760
767                    761
768                    762
769                    763
770                    764
771                    765
772                    766
773                    767
774                    768

-- append local Moran’s I dataframe

Append local Moran’s I dataframe (i.e.localMI) before mapping the local Moran’s I map.

Show the code
nga_wp.localMI <- cbind(nga_wp,localMI) %>%
  rename(Pr.Ii = Pr.z....E.Ii..)

3.7.1.2 plot local Moran’s I values

Show the code
tm_shape(nga_wp.localMI) +
  tm_fill(col = "Ii", 
          style = "pretty",
          palette = "RdBu",
          title = "local moran statistics") +
  tm_borders(alpha = 0.5)
Variable(s) "Ii" contains positive and negative values, so midpoint is set to 0. Set midpoint = NA to show the full spectrum of the color palette.

3.7.1.3 plot local Moran’s I p-value

Show the code
tm_shape(nga_wp.localMI) +
  tm_fill(col = "Pr.Ii", 
          breaks=c(-Inf, 0.001, 0.01, 0.05, 0.1, Inf),
          palette="-Blues", 
          title = "Local Moran's I p-value") +
  tm_borders(alpha = 0.5)

Show the code
localMI.map <- tm_shape(nga_wp.localMI) +
  tm_fill(col = "Ii", 
          style = "pretty",
          palette = "RdBu",
          title = "local moran statistics") +
  tm_borders(alpha = 0.5)  +
  tm_layout(legend.height = 0.3, legend.width = 0.3)


pvalue.map <- tm_shape(nga_wp.localMI) +
  tm_fill(col = "Pr.Ii", 
          breaks=c(-Inf, 0.001, 0.01, 0.05, 0.1, Inf),
          palette="-Blues", 
          title = "local Moran's I p-values") +
  tm_borders(alpha = 0.5) +
  tm_layout(legend.height = 0.26, legend.width = 0.3)

tmap_arrange(localMI.map, pvalue.map, asp=1.5, ncol=1)
Variable(s) "Ii" contains positive and negative values, so midpoint is set to 0. Set midpoint = NA to show the full spectrum of the color palette.

3.8 Creating LISA Cluster Map

Plot the Moran scatterplot before generate the LISA cluster map.

3.8.1 Plot Moran Scatterplot

Use the code chunk below.

The Moran scatterplot is an illustration of the relationship between the values of the chosen attribute at each location and the average value of the same attribute at neighboring locations.

Show the code
nci <- moran.plot(nga_wp$wpt_nonFunctional, ngaWp_rswmQW,
                  labels=as.character(nga_wp$shapeName), 
                  xlab="Non-Functional Water Points", 
                  ylab="Spatially Lag wpt_nonFunctional")

3.8.1.1 plot Moran Scatterplot with Standardised Variable

Show the code
nga_wp$z.wpt_nonFunctional <- scale(nga_wp$wpt_nonFunctional) %>% as.vector 

3.8.1.2 plot Moran Scatterplot

Show the code
nci2 <- moran.plot(nga_wp$z.wpt_nonFunctional, ngaWp_rswmQW,
                   labels=as.character(nga_wp$shapeName),
                   xlab="z-wpt_nonFunctional", 
                   ylab="Spatially Lag z-wpt_nonFunctional")

3.8.2 Prepare LISA Map Classes

Use code chunk below.

Show the code
quadrant <- vector(mode="numeric",length=nrow(localMI))
DV <- nga_wp$wpt_nonFunctional - mean(nga_wp$wpt_nonFunctional)    
C_mI <- localMI[,1] - mean(localMI[,1])    
signif <- 0.05       
quadrant[DV >0 & C_mI>0] <- 4      
quadrant[DV <0 & C_mI<0] <- 1      
quadrant[DV <0 & C_mI>0] <- 2
quadrant[DV >0 & C_mI<0] <- 3
quadrant[localMI[,5]>signif] <- 0

3.8.3 Plot LISA Map

Show the code
nga_wp.localMI$quadrant <- quadrant
colors <- c("#ffffff", "#2c7bb6", "#abd9e9", "#fdae61", "#d7191c")
clusters <- c("insignificant", "low-low", "low-high", "high-low", "high-high")

tm_shape(nga_wp.localMI) +
  tm_fill(col = "quadrant", 
          style = "cat", 
          palette = colors[c(sort(unique(quadrant)))+1], 
          labels = clusters[c(sort(unique(quadrant)))+1],
          popup.vars = c("")) +
  tm_view(set.zoom.limits = c(11,17)) +
  tm_borders(alpha=0.5)

Show the code
wpt_nonFunctional <- qtm(nga_wp, "wpt_nonFunctional") +
  tm_layout(legend.height = 0.28, legend.width = 0.25)

nga_wp.localMI$quadrant <- quadrant 
colors <- c("#ffffff", "#2c7bb6", "#abd9e9", "#fdae61", "#d7191c")
clusters <- c("insignificant", "low-low", "low-high", "high-low", "high-high") 

LISAmap <- tm_shape(nga_wp.localMI) +
  tm_fill(col = "quadrant", 
          style = "cat", 
          palette = colors[c(sort(unique(quadrant)))+1], 
          labels = clusters[c(sort(unique(quadrant)))+1],
          popup.vars = c("")) +
  tm_view(set.zoom.limits = c(11,17)) +
  tm_borders(alpha=0.5) +
  tm_layout(legend.height = 0.35, legend.width = 0.3)

tmap_arrange(wpt_nonFunctional, LISAmap, asp=1.5, ncol=1)


3.9 Hot Spot and Cold Spot Area Analysis

3.9.1 Analyse with Getis and Ord’s G-Statistics

Show the code
k1 <- knn2nb(knearneigh(coords))
k1dists <- unlist(nbdists(k1, coords, longlat = TRUE))
summary(k1dists)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  13.86   28.12   34.99   36.19   44.54   67.50 
Show the code
upper_threshold <- max(unlist(nbdists(k1, coords, longlat = TRUE)))
Show the code
wm_d39 <- dnearneigh(coords, 0, upper_threshold, longlat = TRUE)
wm_d39
Show the code
wm39_lw <- nb2listw(wm_d39, style = 'B')
summary(wm39_lw)
Show the code
knn <- knn2nb(knearneigh(coords, k=8))
knn
Show the code
knn_lw <- nb2listw(knn, style = 'B')
summary(knn_lw)

3.9.2 Compute Gi Statistics Using Fixed Distance

Show the code
fips <- order(nga_wp$shapeName)
gi.fixed <- localG(nga_wp$wpt_nonFunctional, wm39_lw)
gi.fixed
Show the code
nga_wp.gi <- cbind(nga_wp, as.matrix(gi.fixed)) %>%
  rename(gstat_fixed = as.matrix.gi.fixed.)
Show the code
wpt_nonFunctional <- qtm(nga_wp, "wpt_nonFunctional") +
  tm_layout(legend.height = 0.28, legend.width = 0.25)

Gimap <-tm_shape(nga_wp.gi) +
  tm_fill(col = "gstat_fixed", 
          style = "pretty",
          palette="-RdBu",
          title = "local Gi") +
  tm_borders(alpha = 0.5) +
  tm_layout(legend.height = 0.28, legend.width = 0.25)

tmap_arrange(nga_wp, Gimap, asp=1.5, ncol=1)

3.9.3 Compute Gi Statistics Using Adaptive Distance

Show the code
fips <- order(nga_wp$shapeName)
gi.adaptive <- localG(nga_wp$wpt_nonFunctional, knn_lw)
nga_wp.gi <- cbind(nga_wp, as.matrix(gi.adaptive)) %>%
  rename(gstat_adaptive = as.matrix.gi.adaptive.)
Show the code
wpt_nonFunctional <- qtm(nga_wp, "wpt_nonFunctional") +
  tm_layout(legend.height = 0.28, legend.width = 0.25)

Gimap <- tm_shape(nga_wp.gi) + 
  tm_fill(col = "gstat_adaptive", 
          style = "pretty", 
          palette="-RdBu", 
          title = "local Gi") + 
  tm_borders(alpha = 0.5) + 
  tm_layout(legend.height = 0.28, legend.width = 0.25)

tmap_arrange(wpt_nonFunctional,Gimap,asp=1.5,ncol=1)