Home / Questions / 391454
Accepted
Nicx R
Asked: 2020-09-23 08:34:35 +0800 CST 2020-09-23 08:34:35 +0800 CST

Group data into intervals in R

  • 772

I have to organize a thousand data that is in a text file in 10 categories, I have already managed to import the data into R, and create the intervals but I do not know how to organize the data in these intervals, what I need to know is how much data is there in each interval . I hope you can help me.

# establecer el directorio de trabajo
setwd(dir = "C:/Users/DELL/Documents/Texto/")

# importar base de datos en el objeto llamado base_de_datos
base_de_datos <- read.table(file = "taller03_datos.txt")

# extraer los valores de la varible en el objeto llamado cal
cal <- base_de_datos$V1

# tamaño de la muestra
n <- length(cal)

#Máximo y Mínimo
Max <- max(cal)
Min <- min(cal)

#Rango
R <- Max - Min

#promedio aritmético
M <- mean(cal)

#amplitud del intervalo
a <- R/10

#intervalos
li <- seq(from=Min, to= Max-a, by=a)
ls <- seq(from=Min+a, to= Max, by=a)

# Marca de clase
Mc <- (li+ls)/2

# Frecuencia absoluta

#Tabla
tabla <- cbind(li,ls)
r

2 Answers

  1. Best Answer

    If I understand you correctly, I think you have complicated your life, what you are looking for is the frequency for the 10 equidistant intervals. Let's imagine a variable with 1000 observations that can have different values:

    set.seed(2020)
variable <- runif(1000)

    Simply, we can use cut()to "cut" the sample into a desired number of portions, then yes, with table()we can count how many values ​​each portion takes:

    table(cut(variable, 10))

(-0.000105,0.101]     (0.101,0.201]       (0.201,0.3]         (0.3,0.4]         (0.4,0.5] 
              112                93                99                96               115 
        (0.5,0.6]         (0.6,0.7]         (0.7,0.8]       (0.8,0.899]         (0.899,1] 
              107                95                88               104                91
    • 0

  2. To divide a probability distribution into "equal" groups, do we calculate the quantiles, in this case, the deciles? since we want to divide the distribution into 10 parts, which will be very useful for making a box plot for similar groups and observing the outliers.

    quantile(Table$Variable,prob=c(0,0.10,0.20,0.30,0.40,0.50,0.60,0.70,0.80,0.90)

    #on the other hand, if we divide the distribution into 5 parts, we will have the quintiles. Useful for making 5 groups of data.

    quantile(Table$Variable,prob=c(0,0.20,0.4,0.6,0.8,1))

    #Another important grouping, which has to do with sigma values

    quantile(Table$Variable,prob=c(0.6826,0.9546,0.9973,0.999937,0.9999943,0.99999998))

    I don't know if this is exactly what you are looking for.

    • 0