rm(list = ls()) # limpiar el working environment
#carga de data
linkToData='https://docs.google.com/spreadsheets/d/e/2PACX-1vSRpCC8gKIxMxpK0wjgLcl-GQWdw6sAeB16Sixkq6kZXeTfMS8_n70twbEbQ2tMcGp8tm-6x8qf8ieo/pub?gid=234740779&single=true&output=csv'
pavi=read.csv(linkToData)
str(pavi)
## 'data.frame': 1096 obs. of 9 variables:
## $ municipio : chr "m1" "m2" "m3" "m4" ...
## $ apropiaciondolar: num 102.17 4.19 1.59 0 0 ...
## $ consejocomunal : int 0 0 0 0 0 0 0 0 0 0 ...
## $ ejecucion : int 0 0 0 0 0 0 0 0 0 0 ...
## $ uribista : int 0 1 1 1 1 1 1 1 1 NA ...
## $ pctopo : num 14.82 14.51 15.08 6.15 47.31 ...
## $ priorizado : int 0 1 0 0 0 0 0 0 1 0 ...
## $ poblacioncienmil: num 20.9155 0.2406 0.0398 0.0558 0.2723 ...
## $ nbi : num 12.2 33.8 28.5 33.1 27.1 ...
pavi=pavi[complete.cases(pavi),]
seleccion=c("consejocomunal","ejecucion","uribista","priorizado")
pavi[,seleccion]=lapply(pavi[,seleccion],as.factor)
paviStats=summary(pavi[,-1])
paviStats
## apropiaciondolar consejocomunal ejecucion uribista pctopo
## Min. : 0.000 0:827 0:844 0:325 Min. : 0.000
## 1st Qu.: 0.000 1: 49 1: 32 1:551 1st Qu.: 6.285
## Median : 0.000 Median :19.607
## Mean : 8.926 Mean :27.306
## 3rd Qu.: 11.436 3rd Qu.:44.455
## Max. :132.643 Max. :99.419
## priorizado poblacioncienmil nbi
## 0:639 Min. : 0.00629 Min. : 6.84
## 1:237 1st Qu.: 0.07460 1st Qu.:27.33
## Median : 0.14593 Median :40.26
## Mean : 0.45639 Mean :41.72
## 3rd Qu.: 0.27433 3rd Qu.:53.78
## Max. :69.26836 Max. :97.32
library(ggcorrplot)
## Loading required package: ggplot2
colNums=names(pavi)[c(2,6,8,9)]
numXs=pavi[,colNums]
ggcorrplot(cor(numXs),lab = T,show.diag = F)
library(magrittr)
library(kableExtra)
colCats=setdiff(names(pavi), colNums)[-1]
diffPara=c()
diffNoPara=c()
for (col in colCats){
diffPara=c(diffPara,t.test(pavi[,"apropiaciondolar"]~pavi[,col])['p.value']<=0.05)
diffNoPara=c(diffNoPara,wilcox.test(pavi[,"apropiaciondolar"]~pavi[,col])['p.value']<=0.05)
}
data.frame(cbind(colCats,diffPara,diffNoPara),
row.names = 1:length(colCats))%>%
kable(caption = "Diferencia de 'VD:Apropiacion' por Grupo")%>%
kableExtra::kable_styling(full_width = FALSE)
Diferencia de ‘VD:Apropiacion’ por Grupo
|
colCats
|
diffPara
|
diffNoPara
|
|
consejocomunal
|
TRUE
|
TRUE
|
|
ejecucion
|
FALSE
|
FALSE
|
|
uribista
|
TRUE
|
FALSE
|
|
priorizado
|
FALSE
|
TRUE
|
par(mfrow = c(2, 2))
for (col in colCats) {
boxplot(pavi$apropiaciondolar~pavi[,col],
main = col,xlab="",ylab = "")
}
# hipotesis en R
modelo1=formula(apropiaciondolar~pctopo+poblacioncienmil)
reg1=lm(modelo1,data=pavi)
summary(reg1)
##
## Call:
## lm(formula = modelo1, data = pavi)
##
## Residuals:
## Min 1Q Median 3Q Max
## -53.809 -8.671 -7.448 2.602 90.426
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 8.80319 0.79976 11.007 <2e-16 ***
## pctopo -0.03146 0.02150 -1.463 0.144
## poblacioncienmil 2.15125 0.20073 10.717 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 15.85 on 873 degrees of freedom
## Multiple R-squared: 0.1181, Adjusted R-squared: 0.1161
## F-statistic: 58.45 on 2 and 873 DF, p-value: < 2.2e-16
# modelo 1 mas una nueva independiente
modelo2=formula(apropiaciondolar~pctopo+consejocomunal+poblacioncienmil)
reg2=lm(modelo2,data=pavi)
summary(reg2)
##
## Call:
## lm(formula = modelo2, data = pavi)
##
## Residuals:
## Min 1Q Median 3Q Max
## -62.378 -7.933 -6.834 1.573 91.247
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 8.04119 0.79129 10.162 < 2e-16 ***
## pctopo -0.02987 0.02103 -1.420 0.156
## consejocomunal1 14.79601 2.32127 6.374 2.98e-10 ***
## poblacioncienmil 1.91236 0.19987 9.568 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 15.51 on 872 degrees of freedom
## Multiple R-squared: 0.1573, Adjusted R-squared: 0.1545
## F-statistic: 54.28 on 3 and 872 DF, p-value: < 2.2e-16
# modelo 2 mas 'uribista'
modelo3=formula(apropiaciondolar~pctopo+consejocomunal+uribista+poblacioncienmil)
reg3=lm(modelo3,data=pavi)
summary(reg3)
##
## Call:
## lm(formula = modelo3, data = pavi)
##
## Residuals:
## Min 1Q Median 3Q Max
## -61.004 -7.546 -6.482 1.847 91.958
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 9.80542 1.11705 8.778 < 2e-16 ***
## pctopo -0.03753 0.02126 -1.765 0.0779 .
## consejocomunal1 14.73795 2.31613 6.363 3.19e-10 ***
## uribista1 -2.45126 1.09801 -2.232 0.0258 *
## poblacioncienmil 1.89052 0.19965 9.469 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 15.47 on 871 degrees of freedom
## Multiple R-squared: 0.1621, Adjusted R-squared: 0.1583
## F-statistic: 42.14 on 4 and 871 DF, p-value: < 2.2e-16
library(modelsummary)
## `modelsummary` 2.0.0 now uses `tinytable` as its default table-drawing
## backend. Learn more at: https://vincentarelbundock.github.io/tinytable/
##
## Revert to `kableExtra` for one session:
##
## options(modelsummary_factory_default = 'kableExtra')
##
## Change the default backend persistently:
##
## config_modelsummary(factory_default = 'gt')
##
## Silence this message forever:
##
## config_modelsummary(startup_message = FALSE)
modelo3_st=formula(scale(apropiaciondolar)~scale(pctopo)+scale(as.numeric(consejocomunal))+scale(as.numeric(uribista))+scale(poblacioncienmil))
modelo3_st=lm(modelo3_st,data=pavi)
modelo3_st=list('apropiacion (III_st)'=modelo3_st)
modelsummary(modelo3_st, title = "Regresion: modelo 3 con \ncoeficientes estandarizados",
stars = TRUE,
output = "kableExtra")
Regresion: modelo 3 con coeficientes estandarizados
|
|
apropiacion (III_st)
|
|
(Intercept)
|
0.000
|
|
|
(0.031)
|
|
scale(pctopo)
|
-0.055+
|
|
|
(0.031)
|
|
scale(as.numeric(consejocomunal))
|
0.201***
|
|
|
(0.032)
|
|
scale(as.numeric(uribista))
|
-0.070*
|
|
|
(0.031)
|
|
scale(poblacioncienmil)
|
0.299***
|
|
|
(0.032)
|
|
Num.Obs.
|
876
|
|
R2
|
0.162
|
|
R2 Adj.
|
0.158
|
|
AIC
|
2342.0
|
|
BIC
|
2370.7
|
|
Log.Lik.
|
-1165.004
|
|
F
|
42.140
|
|
RMSE
|
0.91
|
|
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p <
0.001
|