library(rio)
enaho <- import("https://github.com/JairAlva/Estadistica-II/raw/master/enaho_2018_m85.sav")str(enaho$P39B1)## num [1:37462] 5 5 6 3 3 4 1 NA 5 7 ...
## - attr(*, "label")= chr "Si la condición económica de su hogar es medida en una escala del 1 al 10 ¿En que escalón considera se encuentr"| __truncated__
## - attr(*, "format.spss")= chr "F2.0"
## - attr(*, "display_width")= int 7str(enaho$P208_02)## num [1:37462] 49 58 47 59 47 30 40 NA 27 61 ...
## - attr(*, "label")= chr "¿Qué edad tiene en años cumplidos?"
## - attr(*, "format.spss")= chr "F2.0"
## - attr(*, "display_width")= int 9sapply(enaho[c("P39B1", "P208_02", "P301_02", "P207_02")], class)## P39B1 P208_02 P301_02 P207_02
## "numeric" "numeric" "numeric" "numeric"modelo1=formula( P39B1~ P208_02+ P301_02)reg1=lm(modelo1,data=enaho)
summary(reg1)##
## Call:
## lm(formula = modelo1, data = enaho)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.8629 -1.0121 -0.2602 0.7638 7.3950
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.0796829 0.0347360 31.083 < 2e-16 ***
## P208_02 0.0023984 0.0004883 4.912 9.08e-07 ***
## P301_02 0.3591330 0.0031478 114.089 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.352 on 33936 degrees of freedom
## (3523 observations deleted due to missingness)
## Multiple R-squared: 0.2918, Adjusted R-squared: 0.2918
## F-statistic: 6993 on 2 and 33936 DF, p-value: < 2.2e-16library(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)model1=list('Condicion economica'=reg1)
modelsummary(model1, title = "Regresion: modelo 1",
stars = TRUE,
output = "kableExtra")| Condicion economica | |
|---|---|
| (Intercept) | 1.080*** |
| (0.035) | |
| P208_02 | 0.002*** |
| (0.000) | |
| P301_02 | 0.359*** |
| (0.003) | |
| Num.Obs. | 33939 |
| R2 | 0.292 |
| R2 Adj. | 0.292 |
| AIC | 116784.2 |
| BIC | 116817.9 |
| Log.Lik. | -58388.104 |
| RMSE | 1.35 |
| + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001 |
modelo2=formula( P39B1~ P208_02+ P301_02+P207_02)reg2=lm(modelo2,data=enaho)
summary(reg2)##
## Call:
## lm(formula = modelo2, data = enaho)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.9889 -1.0187 -0.2101 0.8085 7.3311
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.7654992 0.0447254 17.116 < 2e-16 ***
## P208_02 0.0030943 0.0004914 6.296 3.08e-10 ***
## P301_02 0.3644871 0.0031788 114.661 < 2e-16 ***
## P207_02 0.1654877 0.0148805 11.121 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.349 on 33935 degrees of freedom
## (3523 observations deleted due to missingness)
## Multiple R-squared: 0.2944, Adjusted R-squared: 0.2944
## F-statistic: 4720 on 3 and 33935 DF, p-value: < 2.2e-16library(magrittr)
library(knitr)
tanova=anova(reg1,reg2)
kable(tanova,
caption = "Tabla ANOVA para comparar modelos")%>%kableExtra::kable_styling(full_width = FALSE)| Res.Df | RSS | Df | Sum of Sq | F | Pr(>F) |
|---|---|---|---|---|---|
| 33936 | 62019.89 | NA | NA | NA | NA |
| 33935 | 61794.68 | 1 | 225.217 | 123.6796 | 0 |
plot(reg1,1
)
###homocedasticidad
library(lmtest)## Loading required package: zoo##
## Attaching package: 'zoo'## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numericlibrary(kableExtra)
# null: modelo homocedastico
resBP=bptest(reg1)
data.frame(list('BP'=resBP$statistic,
'df'=resBP$parameter,
"p-value"=resBP$p.value))%>%
kable(caption = resBP$method)%>%kable_styling(full_width = F)| BP | df | p.value | |
|---|---|---|---|
| BP | 170.2904 | 2 | 0 |
###Normalidad de residuos
plot(reg1, 2)
if (length(reg1$residuals) >= 3 & length(reg1$residuals) <= 5000) {
resSW = shapiro.test(reg1$residuals)
data.frame(SW = resSW$statistic,
`p-value` = resSW$p.value) %>%
kable(caption = resSW$method) %>%
kable_styling(full_width = F)
} else {
print("Shapiro-Wilk solo se aplica entre 3 y 5000 observaciones.")
}## [1] "Shapiro-Wilk solo se aplica entre 3 y 5000 observaciones."length(reg1$residuals)## [1] 33939library(DescTools)##
## Attaching package: 'DescTools'## The following objects are masked from 'package:modelsummary':
##
## Format, Mean, Median, N, SD, Var###NO MULTICOLINEALIDAD
VIF(reg1) %>%kable(col.names = "VIF",caption ="Evaluando Multicolinealidad usando VIF (Variance Inflation Factors)" )%>%kable_styling(full_width = F)| VIF | |
|---|---|
| P208_02 | 1.101172 |
| P301_02 | 1.101172 |
plot(reg1)
###Valores influyentes
plot(reg1,5)
checkReg2=as.data.frame(influence.measures(reg1)$is.inf)
checkReg2[checkReg2$cook.d & checkReg2$hat,c('cook.d','hat')]%>%kable(caption = "Valores Influyentes criticos")%>%kable_styling(full_width = F)| cook.d | hat |
|---|---|
| NA | NA |
| :—— | :— |
##PREGUNTA 2
library(rio)
data <- import("https://github.com/JairAlva/Estadistica-II/raw/master/iop_seguridad_2012.sav")library(dplyr)##
## Attaching package: 'dplyr'## The following object is masked from 'package:kableExtra':
##
## group_rows## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, uniondata$pnp <- recode(data$P7,
`1` = 1, # 1 se mantiene como 1
`2` = 1, # 2 se recodifica a 1
`3` = 0, # 3 se recodifica a 0
`4` = 0, # 4 se recodifica a 0
`9` = NA_real_) #9 como datos perdidossapply(data[c("VP1", "NSEGrup", "EDAD", "P2")], class)## VP1 NSEGrup EDAD P2
## "numeric" "numeric" "numeric" "numeric"