- Paquete estadístico (SPSS, STATA, Excel(?))
- Open Source
- Ecosistema de herramientas (RStudio, dplyr)
2/4/2018
¿Qué es R?
Introducción
- Código:
- Para clonarlo desde GIT:
$ git clone https://github.com/rlabuonora/taller_R.git
Interfaz (1)
- Escribo comandos para evaluarlos directamente
Interfaz (2)
- Archivos de texto salvados con código fuente
Consola
- Es la forma más básica de interactuar con R
6*8
## [1] 48
2^6
## [1] 64
- Arriba y abajo para scrollear los comandos
Funciones predefinidas
sqrt(2)
## [1] 1.414214
- Para ver la ayuda de la funcion sqrt:
?sqrt
Variables
- Sirven para almacenar resultados intermedios y organizar los programas:
RaizDeDos <- sqrt(2) HorasPorAnio = 365 * 60
- En Excel y Stata no existen (!)
Ambiente (environment)
- Es el conjunto de objetos disponibles a través de variables
ls()
## [1] "country_data" "counts" "dPost" "dPre" ## [5] "file" "HorasPorAnio" "id" "life_exp" ## [9] "modelo" "modelo_gdp" "MyAbs" "nations" ## [13] "original" "outliers" "RaizDeDos" "t" ## [17] "title" "vec" "who" "who_euro" ## [21] "writing"
Funciones definidas por usuarios
- Es el conjunto de objetos disponibles a través de variables:
MyAbs <- function(x) {
if ( x < 0 ) -x
else x
}
MyAbs(-1)
## [1] 1
MyAbs(2)
## [1] 2
Importar archivo de datos
- Se pueden importar archivos de text, SPSS, Stata, etc.
who <- read.csv("data/WHO.csv")
Estructura
- Un data frame es una lista de vectores
who <- str(who)
## 'data.frame': 194 obs. of 13 variables: ## $ Country : Factor w/ 194 levels "Afghanistan",..: 1 2 3 4 5 6 7 8 9 10 ... ## $ Region : Factor w/ 6 levels "Africa","Americas",..: 3 4 1 4 1 2 2 4 6 4 ... ## $ Population : int 29825 3162 38482 78 20821 89 41087 2969 23050 8464 ... ## $ Under15 : num 47.4 21.3 27.4 15.2 47.6 ... ## $ Over60 : num 3.82 14.93 7.17 22.86 3.84 ... ## $ FertilityRate : num 5.4 1.75 2.83 NA 6.1 2.12 2.2 1.74 1.89 1.44 ... ## $ LifeExpectancy : int 60 74 73 82 51 75 76 71 82 81 ... ## $ ChildMortality : num 98.5 16.7 20 3.2 163.5 ... ## $ CellularSubscribers : num 54.3 96.4 99 75.5 48.4 ... ## $ LiteracyRate : num NA NA NA NA 70.1 99 97.8 99.6 NA NA ... ## $ GNI : num 1140 8820 8310 NA 5230 ... ## $ PrimarySchoolEnrollmentMale : num NA NA 98.2 78.4 93.1 91.1 NA NA 96.9 NA ... ## $ PrimarySchoolEnrollmentFemale: num NA NA 96.4 79.4 78.2 84.5 NA NA 97.5 NA ...
Data frames
- Para acceder a uno usamos el operador $:
who <- read.csv("data/WHO.csv")
head(who$Country) #
## [1] Afghanistan Albania Algeria ## [4] Andorra Angola Antigua and Barbuda ## 194 Levels: Afghanistan Albania Algeria Andorra ... Zimbabwe
heades una función que me muestra los primeros valores de un vector
head(who$Country)
## [1] Afghanistan Albania Algeria ## [4] Andorra Angola Antigua and Barbuda ## 194 Levels: Afghanistan Albania Algeria Andorra ... Zimbabwe
Manejo de archivos
who <- read.csv("data/WHO.csv")
# subset
who_euro <- subset(who, Region=="Europe")
# salvar como csv
write.csv(who_euro, file="data/euro.csv")
Info sobre el data frame
nrow(who)
## [1] 194
nrow(who_euro)
## [1] 53
ncol(who)
## [1] 13
names(who)
## [1] "Country" "Region" ## [3] "Population" "Under15" ## [5] "Over60" "FertilityRate" ## [7] "LifeExpectancy" "ChildMortality" ## [9] "CellularSubscribers" "LiteracyRate" ## [11] "GNI" "PrimarySchoolEnrollmentMale" ## [13] "PrimarySchoolEnrollmentFemale"
Estadísticas descriptivas
summary(who)
## Country Region Population ## Afghanistan : 1 Africa :46 Min. : 1 ## Albania : 1 Americas :35 1st Qu.: 1696 ## Algeria : 1 Eastern Mediterranean:22 Median : 7790 ## Andorra : 1 Europe :53 Mean : 36360 ## Angola : 1 South-East Asia :11 3rd Qu.: 24535 ## Antigua and Barbuda: 1 Western Pacific :27 Max. :1390000 ## (Other) :188 ## Under15 Over60 FertilityRate LifeExpectancy ## Min. :13.12 Min. : 0.81 Min. :1.260 Min. :47.00 ## 1st Qu.:18.72 1st Qu.: 5.20 1st Qu.:1.835 1st Qu.:64.00 ## Median :28.65 Median : 8.53 Median :2.400 Median :72.50 ## Mean :28.73 Mean :11.16 Mean :2.941 Mean :70.01 ## 3rd Qu.:37.75 3rd Qu.:16.69 3rd Qu.:3.905 3rd Qu.:76.00 ## Max. :49.99 Max. :31.92 Max. :7.580 Max. :83.00 ## NA's :11 ## ChildMortality CellularSubscribers LiteracyRate GNI ## Min. : 2.200 Min. : 2.57 Min. :31.10 Min. : 340 ## 1st Qu.: 8.425 1st Qu.: 63.57 1st Qu.:71.60 1st Qu.: 2335 ## Median : 18.600 Median : 97.75 Median :91.80 Median : 7870 ## Mean : 36.149 Mean : 93.64 Mean :83.71 Mean :13321 ## 3rd Qu.: 55.975 3rd Qu.:120.81 3rd Qu.:97.85 3rd Qu.:17558 ## Max. :181.600 Max. :196.41 Max. :99.80 Max. :86440 ## NA's :10 NA's :91 NA's :32 ## PrimarySchoolEnrollmentMale PrimarySchoolEnrollmentFemale ## Min. : 37.20 Min. : 32.50 ## 1st Qu.: 87.70 1st Qu.: 87.30 ## Median : 94.70 Median : 95.10 ## Mean : 90.85 Mean : 89.63 ## 3rd Qu.: 98.10 3rd Qu.: 97.90 ## Max. :100.00 Max. :100.00 ## NA's :93 NA's :93
Vectores
- Son la base de los data frames (un data frame es una lista de vectores (?))
- Muchas funciones trabajan con vectores
- Para crear un vector de números uso la función c:
vec <- c(2, 3, 5, 8, 13) vec
## [1] 2 3 5 8 13
- Accedo por el índice a cada componente
vec[3]
## [1] 5
Vectores (2)
- Pueden ser de texto
counts <- c("Brazil", "China", "India", "Switzerland", "USA")
life_exp<- c(74, 76, 65, 83, 79)
- Para crear un data frame a partir de vectores:
country_data <- data.frame(counts, life_exp) str(country_data)
## 'data.frame': 5 obs. of 2 variables: ## $ counts : Factor w/ 5 levels "Brazil","China",..: 1 2 3 4 5 ## $ life_exp: num 74 76 65 83 79
Vectores (3) -Operador $
- El operador $ me permite acceder a un item de una lista
country_data$pop <- c(199000, 1390000, 1240000, 7997, 318000) str(country_data)
## 'data.frame': 5 obs. of 3 variables: ## $ counts : Factor w/ 5 levels "Brazil","China",..: 1 2 3 4 5 ## $ life_exp: num 74 76 65 83 79 ## $ pop : num 199000 1390000 1240000 7997 318000
- ¡Cuidado con el orden!
Preguntas más específicas sobre los datos
- País con menor proporción de menores de 15
which.min(who$Under15)
## [1] 86
who$Country[which.min(who$Under15)]
## [1] Japan ## 194 Levels: Afghanistan Albania Algeria Andorra ... Zimbabwe
Visualización
- ¿Cuál es la relación entre el PIB y la tasa de fertilidad?
plot(who$GNI, who$FertilityRate)
Subset
Hay dos países que no siguen la relacion negativa:
outliers <- subset(who, GNI > 80000 | (FertilityRate > 5 & GNI > 20000))
outliers[c("Country", "GNI", "FertilityRate")]
## Country GNI FertilityRate ## 56 Equatorial Guinea 25620 5.04 ## 138 Qatar 86440 2.06
Histogramas
hist(who$CellularSubscribers)
Boxplots
boxplot(who$LifeExpectancy ~ who$Region)
boxplot(who$LifeExpectancy ~ who$Region, xlab="", ylab="Life Expectancy", main="Life Expectancy by Region")
Tables
- Si le doy un vector categórico, calcula los totales
table(who$Region)
## ## Africa Americas Eastern Mediterranean ## 46 35 22 ## Europe South-East Asia Western Pacific ## 53 11 27
Tapply
- Le doy un vector continuo, un vector categórico y una función
- Aplica la función a cada grupo del vector categórico
tapply(who$Over60, who$Region, mean)
## Africa Americas Eastern Mediterranean ## 5.220652 10.943714 5.620000 ## Europe South-East Asia Western Pacific ## 19.774906 8.769091 10.162963
tapply(who$LiteracyRate, who$Region, min, na.rm=TRUE)
## Africa Americas Eastern Mediterranean ## 31.1 75.2 63.9 ## Europe South-East Asia Western Pacific ## 95.2 56.8 60.6
Importar funcionalidad
La función library permite importar "bibliotecas"
## install.packages("readstata13")
library(readstata13)
writing <- readstata13::read.dta13("data/writing.dta")
head(writing)
## id preS preP preC preE postS postP postC postE ## 1 A 11 2 1 2 23 5 2 6 ## 2 B 8 1 1 0 9 5 1 1 ## 3 C 11 1 0 0 24 4 1 1 ## 4 D 8 1 0 0 20 6 2 5 ## 5 E 6 1 0 1 16 2 1 2 ## 6 F 8 1 0 2 18 5 1 1
Los datos
Oraciones completas antes y después del tratamientos
- preS
- postS
summary(writing[c("preS", "postS")])
## preS postS ## Min. : 3.00 Min. : 9.00 ## 1st Qu.: 8.00 1st Qu.:21.75 ## Median :10.50 Median :24.50 ## Mean :10.79 Mean :26.38 ## 3rd Qu.:13.00 3rd Qu.:30.75 ## Max. :21.00 Max. :47.00
Gráficamente
boxplot(writing$preS, writing$postS)
Distribución
¿Es normal?
dPre <- density(writing$preS) plot(dPre)
Distribución
¿Es normal?
dPost <- density(writing$postS) plot(dPost)
Estadísticamente
- Prueba de hipótesis:
t.test(writing$postS, writing$preS )
## ## Welch Two Sample t-test ## ## data: writing$postS and writing$preS ## t = 8.0441, df = 35.945, p-value = 1.492e-09 ## alternative hypothesis: true difference in means is not equal to 0 ## 95 percent confidence interval: ## 11.65424 19.51243 ## sample estimates: ## mean of x mean of y ## 26.37500 10.79167
- Nos da un intervalo de confianza para la diferencia de medias.
Modelos lineales
Archivos nativos R
nations <- readRDS("data/nations.rds")
summary(nations)
## country region gdp school ## Length:194 Africa :52 Min. : 279.8 Min. : 1.150 ## Class :character Americas:35 1st Qu.: 2100.3 1st Qu.: 5.225 ## Mode :character Asia :49 Median : 6932.8 Median : 7.950 ## Europe :43 Mean :12118.7 Mean : 7.459 ## Oceania :15 3rd Qu.:17576.3 3rd Qu.: 9.850 ## Max. :74906.0 Max. :12.700 ## NA's :15 NA's :6 ## adfert chldmort life pop ## Min. : 1.00 Min. : 2.25 Min. :45.85 Min. :9.767e+03 ## 1st Qu.: 16.50 1st Qu.: 10.25 1st Qu.:63.75 1st Qu.:1.484e+06 ## Median : 39.35 Median : 24.00 Median :72.18 Median :6.785e+06 ## Mean : 51.81 Mean : 47.65 Mean :68.73 Mean :3.438e+07 ## 3rd Qu.: 74.38 3rd Qu.: 69.75 3rd Qu.:75.55 3rd Qu.:2.226e+07 ## Max. :207.10 Max. :209.00 Max. :82.77 Max. :1.325e+09 ## NA's :1 ## urban femlab literacy co2 ## Min. : 10.25 Min. :0.1942 Min. : 23.60 Min. : 0.10 ## 1st Qu.: 36.33 1st Qu.:0.5782 1st Qu.: 70.10 1st Qu.: 2.25 ## Median : 56.48 Median :0.7304 Median : 89.70 Median : 8.45 ## Mean : 55.43 Mean :0.6916 Mean : 81.53 Mean : 17.89 ## 3rd Qu.: 73.48 3rd Qu.:0.8206 3rd Qu.: 97.70 3rd Qu.: 25.00 ## Max. :100.00 Max. :1.0344 Max. :100.00 Max. :210.65 ## NA's :17 NA's :59 NA's :9 ## gini loggdp reg1 reg2 ## Min. :19.00 Min. :2.447 others:142 Min. :0.0000 ## 1st Qu.:34.00 1st Qu.:3.322 Africa: 52 1st Qu.:0.0000 ## Median :39.70 Median :3.841 Median :0.0000 ## Mean :40.48 Mean :3.776 Mean :0.1804 ## 3rd Qu.:46.20 3rd Qu.:4.245 3rd Qu.:0.0000 ## Max. :58.50 Max. :4.875 Max. :1.0000 ## NA's :113 NA's :15 ## reg3 reg4 reg5 logco2 ## Min. :0.0000 others:151 Min. :0.00000 Min. :-1.0000 ## 1st Qu.:0.0000 Europe: 43 1st Qu.:0.00000 1st Qu.: 0.3522 ## Median :0.0000 Median :0.00000 Median : 0.9269 ## Mean :0.2526 Mean :0.07732 Mean : 0.8069 ## 3rd Qu.:0.7500 3rd Qu.:0.00000 3rd Qu.: 1.3979 ## Max. :1.0000 Max. :1.00000 Max. : 2.3236 ## NA's :9 ## urb_reg4 urban0 loggdp0 ## Min. : 0.00 Min. :-45.15000 Min. :-1.333152 ## 1st Qu.: 0.00 1st Qu.:-19.07500 1st Qu.:-0.458354 ## Median : 0.00 Median : 1.07500 Median : 0.060909 ## Mean : 15.53 Mean : 0.03488 Mean :-0.004271 ## 3rd Qu.: 0.00 3rd Qu.: 18.08333 3rd Qu.: 0.464920 ## Max. :100.00 Max. : 44.60000 Max. : 1.094517 ## NA's :15 ## urb_gdp ## Min. :-24.4014 ## 1st Qu.: 0.6885 ## Median : 6.1521 ## Mean : 9.6186 ## 3rd Qu.: 17.2818 ## Max. : 55.8472 ## NA's :15
Explicamos la variación en esperanza de vida
Un modelo lineal:
modelo_gdp <- lm(life ~ school, data=nations) summary(modelo_gdp)
## ## Call: ## lm(formula = life ~ school, data = nations) ## ## Residuals: ## Min 1Q Median 3Q Max ## -20.650 -4.058 1.529 4.747 14.393 ## ## Coefficients: ## Estimate Std. Error t value Pr(>|t|) ## (Intercept) 50.3594 1.3692 36.78 <2e-16 *** ## school 2.4518 0.1707 14.37 <2e-16 *** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Residual standard error: 6.908 on 186 degrees of freedom ## (6 observations deleted due to missingness) ## Multiple R-squared: 0.5259, Adjusted R-squared: 0.5234 ## F-statistic: 206.3 on 1 and 186 DF, p-value: < 2.2e-16
Visualizamos la relación entre variables
Gráficamente:
Modelo
- Especificamos un modelo multivariado para la esperanza de vida:
\[life= \beta_0 + \beta_1 * school + \beta_2 * log(gdp) + \beta_3 * adfert + \beta_4 childmort\]
Para estimarlo
modelo <- lm(life ~ school + log(gdp) + adfert + chldmort, data=nations) summary(modelo)
## ## Call: ## lm(formula = life ~ school + log(gdp) + adfert + chldmort, data = nations) ## ## Residuals: ## Min 1Q Median 3Q Max ## -15.4363 -1.6753 0.3561 2.2653 6.3244 ## ## Coefficients: ## Estimate Std. Error t value Pr(>|t|) ## (Intercept) 62.2544019 3.1144342 19.989 < 2e-16 *** ## school -0.2339704 0.1558288 -1.501 0.135 ## log(gdp) 1.7601685 0.3532287 4.983 1.51e-06 *** ## adfert -0.0004683 0.0096839 -0.048 0.961 ## chldmort -0.1511827 0.0098966 -15.276 < 2e-16 *** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Residual standard error: 3.478 on 173 degrees of freedom ## (16 observations deleted due to missingness) ## Multiple R-squared: 0.8813, Adjusted R-squared: 0.8786 ## F-statistic: 321.2 on 4 and 173 DF, p-value: < 2.2e-16
Conclusiones
- Si controlamos por el ingreso, la educación no afecta la esperanza de vida
- La relación entre la esperanza de vida y el ingreso no es lineal, pero es positiva
Deberes
La propuesta esta en la carpeta demo (demo_ej.r)
La solución es (demo_sol.r)
Requiere usar varias funciones:
- str
- is.na
- sort
- names
- length
- table
- tapply