DEM 7283 Stats II Homework 4
Assogba Albert
3/2/2021
Load libraries
library(car)## Loading required package: carDatalibrary(stargazer)##
## Please cite as:## Hlavac, Marek (2018). stargazer: Well-Formatted Regression and Summary Statistics Tables.## R package version 5.2.2. https://CRAN.R-project.org/package=stargazerlibrary(survey)## Loading required package: grid## Loading required package: Matrix## Loading required package: survival##
## Attaching package: 'survey'## The following object is masked from 'package:graphics':
##
## dotchartlibrary(questionr)
library(dplyr)##
## Attaching package: 'dplyr'## The following object is masked from 'package:car':
##
## recode## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(tidyverse)## -- Attaching packages --------------------------------------- tidyverse 1.3.0 --## v ggplot2 3.3.3 v purrr 0.3.4
## v tibble 3.1.0 v stringr 1.4.0
## v tidyr 1.1.2 v forcats 0.5.1
## v readr 1.4.0## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x tidyr::expand() masks Matrix::expand()
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
## x tidyr::pack() masks Matrix::pack()
## x dplyr::recode() masks car::recode()
## x purrr::some() masks car::some()
## x tidyr::unpack() masks Matrix::unpack()library(haven)
library(readr)
library(tableone, quietly = T)
library(ggplot2, quietly = T)
library(caret, quietly = T)##
## Attaching package: 'caret'## The following object is masked from 'package:purrr':
##
## lift## The following object is masked from 'package:survival':
##
## clusterLoad data
GSS2018 <- read_csv("C:/Users/chris/OneDrive/Desktop/GSS2018/GSS2018.csv")##
## -- Column specification --------------------------------------------------------
## cols(
## .default = col_double()
## )
## i Use `spec()` for the full column specifications.# Work on names of variables
nams<- names(GSS2018)
newnames<-tolower(gsub(pattern = "_",replacement = "",x = nams))
names(GSS2018)<-newnames
head(GSS2018)## # A tibble: 6 x 1,067
## year id wrkstat hrs1 hrs2 evwork wrkslf wrkgovt occ10 prestg10
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 2018 1 3 -1 41 0 2 2 630 47
## 2 2018 2 5 -1 -1 1 2 2 9640 22
## 3 2018 3 1 40 -1 0 2 2 1106 61
## 4 2018 4 1 40 -1 0 2 2 3320 59
## 5 2018 5 5 -1 -1 1 2 2 136 53
## 6 2018 6 5 -1 -1 1 2 2 120 53
## # ... with 1,057 more variables: prestg105plus <dbl>, indus10 <dbl>,
## # marital <dbl>, martype <dbl>, divorce <dbl>, widowed <dbl>, spwrksta <dbl>,
## # sphrs1 <dbl>, sphrs2 <dbl>, spevwork <dbl>, cowrksta <dbl>, cowrkslf <dbl>,
## # coevwork <dbl>, cohrs1 <dbl>, cohrs2 <dbl>, spwrkslf <dbl>, spocc10 <dbl>,
## # sppres10 <dbl>, sppres105plus <dbl>, spind10 <dbl>, coocc10 <dbl>,
## # coind10 <dbl>, pawrkslf <dbl>, paocc10 <dbl>, papres10 <dbl>,
## # papres105plus <dbl>, paind10 <dbl>, mawrkslf <dbl>, maocc10 <dbl>,
## # mapres10 <dbl>, mapres105plus <dbl>, maind10 <dbl>, sibs <dbl>,
## # childs <dbl>, age <dbl>, agekdbrn <dbl>, educ <dbl>, paeduc <dbl>,
## # maeduc <dbl>, speduc <dbl>, coeduc <dbl>, codeg <dbl>, degree <dbl>,
## # padeg <dbl>, madeg <dbl>, spdeg <dbl>, major1 <dbl>, major2 <dbl>,
## # dipged <dbl>, sex <dbl>, race <dbl>, res16 <dbl>, reg16 <dbl>,
## # mobile16 <dbl>, family16 <dbl>, famdif16 <dbl>, mawrkgrw <dbl>,
## # incom16 <dbl>, born <dbl>, parborn <dbl>, granborn <dbl>, hompop <dbl>,
## # babies <dbl>, preteen <dbl>, teens <dbl>, adults <dbl>, unrelat <dbl>,
## # earnrs <dbl>, income <dbl>, rincome <dbl>, income16 <dbl>, rincom16 <dbl>,
## # region <dbl>, xnorcsiz <dbl>, srcbelt <dbl>, size <dbl>, partyid <dbl>,
## # vote12 <dbl>, pres12 <dbl>, if12who <dbl>, vote16 <dbl>, pres16 <dbl>,
## # if16who <dbl>, polviews <dbl>, natspac <dbl>, natenvir <dbl>,
## # natheal <dbl>, natcity <dbl>, natcrime <dbl>, natdrug <dbl>, nateduc <dbl>,
## # natrace <dbl>, natarms <dbl>, nataid <dbl>, natfare <dbl>, natroad <dbl>,
## # natsoc <dbl>, natmass <dbl>, natpark <dbl>, natchld <dbl>, ...View(GSS2018)1) Define a binary outcome of your choosing
I will use US presidential candidate voted for in 2016 election (Clinton or Trump) as my outcome.
2) Fit a predictive logistic regression model using as many predictor variables as you think you need.
Recode Variables
# Sex
GSS2018$male<-as.factor(ifelse(GSS2018$sex==1, "Male", "Female"))
#race/ethnicity
GSS2018$black<-Recode(GSS2018$race, recodes="2=1; else=0")
GSS2018$white<-Recode(GSS2018$race, recodes="1=1; else=0")
GSS2018$other<-Recode(GSS2018$race, recodes="3=1; else=0")
GSS2018$race_eth<-Recode(GSS2018$race, recodes="1='nhwhite'; 2='nhblack'; 3='nhother'", as.factor = T)
#education level
GSS2018$educ<-Recode(GSS2018$educ, recodes="1:4='0Prim'; 5:10='1somehs'; 11:12='2hsgrad'; 13:14='3somecol'; 15:20='4colgrad';else=NA", as.factor=T)
#marital status
GSS2018$marst<-Recode(GSS2018$marital, recodes="1='married'; 3='divorced'; 2='widowed'; 4='separated'; 5='nm'; else=NA", as.factor=T)
# Political affiliation
GSS2018$partyaff<- Recode(GSS2018$partyid, recode= "0=0; 6= 1; else=NA", as.factor=T)
# Political views
GSS2018$polviews<- Recode(GSS2018$polviews, recode= "1:3= 'Liberal'; 5:7= 'Conservative'; else=NA", as.factor=T)
# Presidential candidate voted for in 2016 election
GSS2018$vote<- Recode(GSS2018$pres16, recode= "1=0; 2=1; else=NA", as.factor=T)Survey Design
options(survey.lonely.psu = "adjust")
des<-svydesign(ids= ~1, strata= ~vstrat, weights= ~wtssnr, data = GSS2018 )Logit model
fit.logit<-svyglm(vote ~ race_eth + polviews, design = des, family = binomial)## Warning in eval(family$initialize): non-integer #successes in a binomial glm!summary(fit.logit)##
## Call:
## svyglm(formula = vote ~ race_eth + polviews, design = des, family = binomial)
##
## Survey design:
## svydesign(ids = ~1, strata = ~vstrat, weights = ~wtssnr, data = GSS2018)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -2.4218 0.5622 -4.307 1.85e-05 ***
## race_ethnhother 3.3188 0.7596 4.369 1.41e-05 ***
## race_ethnhwhite 4.5409 0.6017 7.546 1.20e-13 ***
## polviewsLiberal -3.9015 0.2709 -14.400 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1.695266)
##
## Number of Fisher Scoring iterations: 6Creating training and test data set
set.seed(1115)
train<-createDataPartition(y = GSS2018$pres16, p = .80, list=F)
GSS2018train<- GSS2018[train,]
GSS2018test<- GSS2018[-train,]
table(GSS2018train$pres16)##
## 0 1 2 3 4 8 9
## 651 609 462 70 17 27 43wat<- GSS2018 %>%
select(vote, educ, male, black, white, other, marst, partyid, polviews, pres16, partyid) %>%
mutate(educ=as.numeric(educ), vote= as.factor(ifelse(vote=="Clinton", 1, ifelse(vote=="Trump", 0, NA))))library(caret)set.seed(1115)
train<-createDataPartition(wat$pres16, p = .80, list=F)
wattrain<-wat[train,]
wattest<-wat[-train,]
table(wattrain$pres16)##
## 0 1 2 3 4 8 9
## 651 609 462 70 17 27 43prop.table(table(wattrain$pres16))##
## 0 1 2 3 4 8
## 0.346460883 0.324108568 0.245875466 0.037253858 0.009047366 0.014369345
## 9
## 0.022884513summary(wattrain)## vote educ male black white
## NA's:1879 Min. :1.000 Female:1035 Min. :0.0000 Min. :0.0000
## 1st Qu.:3.000 Male : 844 1st Qu.:0.0000 1st Qu.:0.0000
## Median :4.000 Median :0.0000 Median :1.0000
## Mean :3.879 Mean :0.1676 Mean :0.7174
## 3rd Qu.:5.000 3rd Qu.:0.0000 3rd Qu.:1.0000
## Max. :5.000 Max. :1.0000 Max. :1.0000
## NA's :7
## other marst partyid polviews
## Min. :0.000 divorced :319 Min. :0.000 Conservative:589
## 1st Qu.:0.000 married :806 1st Qu.:1.000 Liberal :538
## Median :0.000 nm :531 Median :3.000 NA's :752
## Mean :0.115 separated: 63 Mean :2.982
## 3rd Qu.:0.000 widowed :159 3rd Qu.:5.000
## Max. :1.000 NA's : 1 Max. :9.000
##
## pres16
## Min. :0.000
## 1st Qu.:0.000
## Median :1.000
## Mean :1.285
## 3rd Qu.:2.000
## Max. :9.000
## 3) Report the % correct classification from the training data using the .5 decision rule and again useing the mean of the outcome decision rule
glm1<-glm(as.factor(pres16)~factor(educ)+factor(marst)+factor(polviews)+factor(male)+factor(partyid),
data=wattrain[,-1],
family = binomial)
summary(glm1)##
## Call:
## glm(formula = as.factor(pres16) ~ factor(educ) + factor(marst) +
## factor(polviews) + factor(male) + factor(partyid), family = binomial,
## data = wattrain[, -1])
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.6184 -0.6776 0.5011 0.7563 2.1479
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.00202 0.90025 0.002 0.998210
## factor(educ)2 0.96579 0.88413 1.092 0.274674
## factor(educ)3 1.94323 0.85288 2.278 0.022701 *
## factor(educ)4 2.18965 0.85653 2.556 0.010575 *
## factor(educ)5 3.07658 0.85036 3.618 0.000297 ***
## factor(marst)married -0.32922 0.23195 -1.419 0.155783
## factor(marst)nm -1.11780 0.23976 -4.662 3.13e-06 ***
## factor(marst)separated -0.25940 0.45989 -0.564 0.572724
## factor(marst)widowed 0.11368 0.34600 0.329 0.742496
## factor(polviews)Liberal -0.16616 0.18787 -0.884 0.376461
## factor(male)Male 0.24992 0.15374 1.626 0.104041
## factor(partyid)1 -1.12222 0.27060 -4.147 3.37e-05 ***
## factor(partyid)2 -1.28817 0.27892 -4.618 3.87e-06 ***
## factor(partyid)3 -2.67415 0.31502 -8.489 < 2e-16 ***
## factor(partyid)4 -1.37196 0.31221 -4.394 1.11e-05 ***
## factor(partyid)5 -0.86037 0.31599 -2.723 0.006473 **
## factor(partyid)6 -0.04729 0.33646 -0.141 0.888227
## factor(partyid)7 -0.94516 0.44900 -2.105 0.035287 *
## factor(partyid)9 -1.71126 1.02109 -1.676 0.093756 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1353.0 on 1122 degrees of freedom
## Residual deviance: 1085.4 on 1104 degrees of freedom
## (756 observations deleted due to missingness)
## AIC: 1123.4
##
## Number of Fisher Scoring iterations: 43a) Does changing the decision rule threshold affect your classification accuracy?
tr_pred<- predict(glm1,
newdata = wattrain,
type = "response")
head(tr_pred)## 1 2 3 4 5 6
## 0.6138245 0.8358041 NA 0.9637792 0.8504264 0.5540377Using decision rule
tr_predcl<- as.factor(ifelse(tr_pred>.5, 1, 0))
library(ggplot2)
pred1<-data.frame(pr=tr_pred, gr=tr_predcl, modcon=wattrain$pres16)
pred1%>%
ggplot()+
geom_histogram(aes(x=pr, color=gr, fill=gr))+
ggtitle(label = "Probability of voting for a candidate", subtitle = "Threshold = .5")+
geom_vline(xintercept=.5)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 756 rows containing non-finite values (stat_bin).
pred1%>%
ggplot()+
geom_histogram(aes(x=pr, color=modcon, fill=modcon))+
ggtitle(label = "Probability of voting for a candidate", subtitle = "Truth")+
geom_vline(xintercept=.5)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 756 rows containing non-finite values (stat_bin).
table( tr_predcl,wattrain$pres16)##
## tr_predcl 0 1 2 3 4 8 9
## 0 140 35 12 4 0 1 2
## 1 186 343 330 35 8 13 14wat$pres16=as.factor(wat$pres16)cm1<-confusionMatrix(tr_predcl, as.factor(wattrain$pres16) )## Warning in levels(reference) != levels(data): longer object length is not a
## multiple of shorter object length## Warning in confusionMatrix.default(tr_predcl, as.factor(wattrain$pres16)):
## Levels are not in the same order for reference and data. Refactoring data to
## match.cm1## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1 2 3 4 8 9
## 0 140 35 12 4 0 1 2
## 1 186 343 330 35 8 13 14
## 2 0 0 0 0 0 0 0
## 3 0 0 0 0 0 0 0
## 4 0 0 0 0 0 0 0
## 8 0 0 0 0 0 0 0
## 9 0 0 0 0 0 0 0
##
## Overall Statistics
##
## Accuracy : 0.4301
## 95% CI : (0.4009, 0.4597)
## No Information Rate : 0.3366
## P-Value [Acc > NIR] : 4.617e-11
##
## Kappa : 0.1512
##
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: 0 Class: 1 Class: 2 Class: 3 Class: 4 Class: 8
## Sensitivity 0.4294 0.9074 0.0000 0.00000 0.000000 0.00000
## Specificity 0.9322 0.2134 1.0000 1.00000 1.000000 1.00000
## Pos Pred Value 0.7216 0.3692 NaN NaN NaN NaN
## Neg Pred Value 0.7998 0.8196 0.6955 0.96527 0.992876 0.98753
## Prevalence 0.2903 0.3366 0.3045 0.03473 0.007124 0.01247
## Detection Rate 0.1247 0.3054 0.0000 0.00000 0.000000 0.00000
## Detection Prevalence 0.1728 0.8272 0.0000 0.00000 0.000000 0.00000
## Balanced Accuracy 0.6808 0.5604 0.5000 0.50000 0.500000 0.50000
## Class: 9
## Sensitivity 0.00000
## Specificity 1.00000
## Pos Pred Value NaN
## Neg Pred Value 0.98575
## Prevalence 0.01425
## Detection Rate 0.00000
## Detection Prevalence 0.00000
## Balanced Accuracy 0.50000The accuracy in the train data is 43% accurate which is not too good for prediction of an election winner, 43% sensitivity with a specificity of 93% who vote.
Data prediction useing the mean outcome decision rule
pred_test<-predict(glm1,
newdata=wattest,
type="response")
pred_cl<-factor(ifelse(pred_test > mean( I(wattest$pres16==1)), 1, 0))
table(wattest$pres16,pred_cl)## pred_cl
## 0 1
## 0 18 59
## 1 2 89
## 2 2 79
## 3 1 7
## 4 0 1
## 8 1 3
## 9 2 1confusionMatrix(data = pred_cl,as.factor(wattest$pres16) )## Warning in levels(reference) != levels(data): longer object length is not a
## multiple of shorter object length## Warning in confusionMatrix.default(data = pred_cl, as.factor(wattest$pres16)):
## Levels are not in the same order for reference and data. Refactoring data to
## match.## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1 2 3 4 8 9
## 0 18 2 2 1 0 1 2
## 1 59 89 79 7 1 3 1
## 2 0 0 0 0 0 0 0
## 3 0 0 0 0 0 0 0
## 4 0 0 0 0 0 0 0
## 8 0 0 0 0 0 0 0
## 9 0 0 0 0 0 0 0
##
## Overall Statistics
##
## Accuracy : 0.4038
## 95% CI : (0.3442, 0.4655)
## No Information Rate : 0.3434
## P-Value [Acc > NIR] : 0.02344
##
## Kappa : 0.0991
##
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: 0 Class: 1 Class: 2 Class: 3 Class: 4 Class: 8
## Sensitivity 0.23377 0.9780 0.0000 0.00000 0.000000 0.00000
## Specificity 0.95745 0.1379 1.0000 1.00000 1.000000 1.00000
## Pos Pred Value 0.69231 0.3724 NaN NaN NaN NaN
## Neg Pred Value 0.75314 0.9231 0.6943 0.96981 0.996226 0.98491
## Prevalence 0.29057 0.3434 0.3057 0.03019 0.003774 0.01509
## Detection Rate 0.06792 0.3358 0.0000 0.00000 0.000000 0.00000
## Detection Prevalence 0.09811 0.9019 0.0000 0.00000 0.000000 0.00000
## Balanced Accuracy 0.59561 0.5580 0.5000 0.50000 0.500000 0.50000
## Class: 9
## Sensitivity 0.00000
## Specificity 1.00000
## Pos Pred Value NaN
## Neg Pred Value 0.98868
## Prevalence 0.01132
## Detection Rate 0.00000
## Detection Prevalence 0.00000
## Balanced Accuracy 0.50000tr_predcl<-factor(ifelse(tr_pred>mean(I(wattrain$pres16==1)), 1, 0)) #mean of response
pred2<-data.frame(pr=tr_pred, gr=tr_predcl, modcon=wattrain$pres16)
pred2%>%
ggplot(aes(x=pr, fill=gr))+
geom_histogram(position="identity", alpha=.2)+
ggtitle(label = "Probability of Voting", subtitle = "Threshold = Mean")+
geom_vline(xintercept=mean(I(wattrain$pres16==1)))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 756 rows containing non-finite values (stat_bin).
pred2%>%
ggplot(aes(x=pr, fill=modcon))+
geom_histogram(position="identity", alpha=.2)+
ggtitle(label = "Probability of Voting", subtitle = "Truth")+
geom_vline(xintercept=mean(I(wattrain$pres16==1)))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 756 rows containing non-finite values (stat_bin).
confusionMatrix(tr_predcl,
as.factor(wattrain$pres16),
positive = "1" )## Warning in levels(reference) != levels(data): longer object length is not a
## multiple of shorter object length## Warning in confusionMatrix.default(tr_predcl, as.factor(wattrain$pres16), :
## Levels are not in the same order for reference and data. Refactoring data to
## match.## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1 2 3 4 8 9
## 0 72 11 1 2 0 0 2
## 1 254 367 341 37 8 14 14
## 2 0 0 0 0 0 0 0
## 3 0 0 0 0 0 0 0
## 4 0 0 0 0 0 0 0
## 8 0 0 0 0 0 0 0
## 9 0 0 0 0 0 0 0
##
## Overall Statistics
##
## Accuracy : 0.3909
## 95% CI : (0.3623, 0.4202)
## No Information Rate : 0.3366
## P-Value [Acc > NIR] : 7.878e-05
##
## Kappa : 0.0869
##
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: 0 Class: 1 Class: 2 Class: 3 Class: 4 Class: 8
## Sensitivity 0.22086 0.9709 0.0000 0.00000 0.000000 0.00000
## Specificity 0.97992 0.1034 1.0000 1.00000 1.000000 1.00000
## Pos Pred Value 0.81818 0.3546 NaN NaN NaN NaN
## Neg Pred Value 0.75459 0.8750 0.6955 0.96527 0.992876 0.98753
## Prevalence 0.29029 0.3366 0.3045 0.03473 0.007124 0.01247
## Detection Rate 0.06411 0.3268 0.0000 0.00000 0.000000 0.00000
## Detection Prevalence 0.07836 0.9216 0.0000 0.00000 0.000000 0.00000
## Balanced Accuracy 0.60039 0.5371 0.5000 0.50000 0.500000 0.50000
## Class: 9
## Sensitivity 0.00000
## Specificity 1.00000
## Pos Pred Value NaN
## Neg Pred Value 0.98575
## Prevalence 0.01425
## Detection Rate 0.00000
## Detection Prevalence 0.00000
## Balanced Accuracy 0.50000Chaanging the rule from 0.5 outcome to decision to mean outcome decision resulted in lower percent correct classification but an increase in specificity and lower percentage in the sensitivity.
pred_test<-predict(glm1,
newdata=wattest,
type="response")
pred_cl<-factor(ifelse(pred_test > mean( I(wattest$pres16==1)), 1, 0))
table(wattest$pres16,pred_cl)## pred_cl
## 0 1
## 0 18 59
## 1 2 89
## 2 2 79
## 3 1 7
## 4 0 1
## 8 1 3
## 9 2 1confusionMatrix(data = pred_cl,as.factor(wattest$pres16))## Warning in levels(reference) != levels(data): longer object length is not a
## multiple of shorter object length## Warning in confusionMatrix.default(data = pred_cl, as.factor(wattest$pres16)):
## Levels are not in the same order for reference and data. Refactoring data to
## match.## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1 2 3 4 8 9
## 0 18 2 2 1 0 1 2
## 1 59 89 79 7 1 3 1
## 2 0 0 0 0 0 0 0
## 3 0 0 0 0 0 0 0
## 4 0 0 0 0 0 0 0
## 8 0 0 0 0 0 0 0
## 9 0 0 0 0 0 0 0
##
## Overall Statistics
##
## Accuracy : 0.4038
## 95% CI : (0.3442, 0.4655)
## No Information Rate : 0.3434
## P-Value [Acc > NIR] : 0.02344
##
## Kappa : 0.0991
##
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: 0 Class: 1 Class: 2 Class: 3 Class: 4 Class: 8
## Sensitivity 0.23377 0.9780 0.0000 0.00000 0.000000 0.00000
## Specificity 0.95745 0.1379 1.0000 1.00000 1.000000 1.00000
## Pos Pred Value 0.69231 0.3724 NaN NaN NaN NaN
## Neg Pred Value 0.75314 0.9231 0.6943 0.96981 0.996226 0.98491
## Prevalence 0.29057 0.3434 0.3057 0.03019 0.003774 0.01509
## Detection Rate 0.06792 0.3358 0.0000 0.00000 0.000000 0.00000
## Detection Prevalence 0.09811 0.9019 0.0000 0.00000 0.000000 0.00000
## Balanced Accuracy 0.59561 0.5580 0.5000 0.50000 0.500000 0.50000
## Class: 9
## Sensitivity 0.00000
## Specificity 1.00000
## Pos Pred Value NaN
## Neg Pred Value 0.98868
## Prevalence 0.01132
## Detection Rate 0.00000
## Detection Prevalence 0.00000
## Balanced Accuracy 0.50000