library(effects)
x1 <- predictorEffect("x6", model)
plot(x1)
x2 <- predictorEffect("x2", model)
plot(x2, main="model COVID 19",xlab="X2", ylab="x") #single plot
plot(allEffects(model)) #all plot
library(effects)
x1 <- predictorEffect("x6", model)
plot(x1)
x2 <- predictorEffect("x2", model)
plot(x2, main="model COVID 19",xlab="X2", ylab="x") #single plot
plot(allEffects(model)) #all plot
# check that model is good fit or not
with(model, cbind(res.deviance = deviance, df = df.residual,
p = pchisq(deviance, df.residual, lower.tail=FALSE)))
## odds ratios and 95% CI ***********
exp(cbind(OR = coef(model), confint(model)))
# MORE SUMMARIES ###########################################
anova(model) # Coefficients w/inferential tests
coef(model) # Coefficients
hist(coef(model))
confint(model) # CI for coefficients
hist(confint(model))
resid(model) # Residuals case-by-case
hist(residuals(model),main="model COVID 19" ) # Histogram of residuals
plot(residuals(model), main="model COVID 19" )
logLik(model)
BIC(model)
PseudoR2(model)
predict(model)
hist(predict(model))
#peseudo r square
model$null.deviance
model$deviance
modelChi <- model$null.deviance - model$deviance
pseudo <- modelChi / model$null.deviance
pseudo
# Compute the pseudo p-value
Chidf <- model$df.null - model$df.residual
modelChi <- model$null.deviance - model$deviance
1 - pchisq(modelChi, Chidf)
#RSS(residual sum of square)
RSS <- c(crossprod(model$residuals))
RSS
#Mean square error
MSE <- RSS / length(model$residuals)
MSE
#Root MSE
RMSE <- sqrt(MSE)
RMSE
#Pearson estimated residual variance
sig2 <- RSS / model$df.residual
sig2
library(PLNmodels)
library(ggplot2)
library(corrplot)
Y=read.csv(choose.files())
X=read.csv(choose.files())
c=list(Y,X)
names(c) <- c("output", "input")
d <- prepare_data(c$output, c$input)
model <- PLN(Abundance ~
LANE_WIDTH +
LENGTH +
LOG_AVG_AADT +
LOG_PAVEMENT_CONDITION +
HIGH_FREQ_TRANSIT +
eightyfivepercent_SPEED_ADJ +
P_SPEEDING_ADJ +
SL +
NL, data = d)
print(model) #variational lower bound of the ICL
coef(model) #mu -vectors of means of the latent variable
sigma(model) #covariance matrix of the latent variable
vcov(model) #Variance-Covariance Matrix
fitted(model)
standard_error(model)
plot(fitted(model))
barplot(fitted(model))
#2 tailed z test
z<- coef(model)/standard_error(model)
p<- (1-pnorm(abs(z),0,1))*2
p #pvalue
#2 tailed z test
z<- coef(model)/standard_error(model)
p<- (1-pnorm(abs(z),0,1))*2
p #pvalue