Organisation of the x-axis

Question

I made a histogram in the R but I noticed that the axis has an error, it orders -0 as being greater than -2 and does not follow the sequence, as follows in the picture.

I tried to filter it in the database I generated but when I classify it from the smallest number to the largest it continues like this, have you ever been through something like this? Can you help me?

library("reader")
library("lubridate") 
library("zoo") 
install.packages("dplyr")
install.packages("data.table")
install.packages("ggplot2")
install.packages("zoo")

setwd("C:/Users/Giovanni/Desktop")

dados_ibov <- read.csv('treinosr/Ibovespa_datah.csv',header = TRUE, sep = ",")

data.frame(dados_ibov)

freq <- table(dados_ibov$Var.)

barplot(freq, ylab="Frequência", xlab = "Retornos diários do Ibovespa")

as requested:

dput(head(freq, 20))
structure(c(`-0,00%` = 4L, `-0,01%` = 9L, `-0,02%` = 6L, `-0,03%` = 4L, 
`-0,04%` = 9L, `-0,05%` = 9L, `-0,06%` = 10L, `-0,07%` = 6L, 
`-0,08%` = 13L, `-0,09%` = 4L, `-0,10%` = 8L, `-0,11%` = 9L, 
`-0,12%` = 6L, `-0,13%` = 9L, `-0,14%` = 11L, `-0,15%` = 4L, 
`-0,16%` = 8L, `-0,17%` = 6L, `-0,18%` = 10L, `-0,19%` = 5L), 
  .Dim = 20L, .Dimnames = list(
    dado = c("-0,00%", "-0,01%", "-0,02%", "-0,03%", "-0,04%", 
    "-0,05%", "-0,06%", "-0,07%", "-0,08%", "-0,09%", "-0,10%", 
    "-0,11%", "-0,12%", "-0,13%", "-0,14%", "-0,15%", "-0,16%", 
    "-0,17%", "-0,18%", "-0,19%")), class = "table")

Answer 1

The problem is that you’re creating a graph that has a text variable for what should be a numeric variable. And it can’t be any different as you rely on the "name" of the variables, which will always be a text.

It is possible to solve the specific problem presented above by inverting the vector with the function rev().

barplot(rev(sopt), ylab="Frequência", xlab = "Retornos diários do Ibovespa")

But this solution is no guarantee of properly solving the problem, if we add other elements out of order (-2% before -1%) in the vector freq, the result would be different from expected.

barplot(rev(c(freq, c("-2%" = 3, "-1%" = 5))), 
        ylab="Frequência", xlab = "Retornos diários do Ibovespa")

What we need to do is determine the order of the data that will be drawn according to its name (and there should be a numerical ordering).

percentual <- as.numeric(sub(',', '.', sub('%', '', names(freq))))

# Ou, usando readr
percentual <- readr::parse_number(
  names(freq), locale = readr::locale(decimal_mark = ",")
)

ordem <- order(percentual)
barplot(freq[ordem])

Note that this new form is robust to messy vectors and does not suffer from the problems of the second figure (-1% "smaller" than -2%):

freq2 <- c(freq, c("-2%" = 3, "-1%" = 5))
percentual2 <- readr::parse_number(
  names(freq2), locale = readr::locale(decimal_mark = ",")
)
ordem2 <- order(percentual2)
barplot(freq2[ordem2])

The ggplot2

That said, I think it’s much better to trust data.frames and in the graphics system of ggplot2.

# repetir o percentual pelo numero de vezes que aparece
# engenharia reversa do "table" da pergunta
variacoes <- rep(percentual, freq)

df <- data.frame(var = variacoes)

head(df)

#>     var
#> 1  0.00
#> 2  0.00
#> 3  0.00
#> 4  0.00
#> 5 -0.01
#> 6 -0.01

And then with this data.frame on hand, which should be similar to your data.frame original, just use the geom_bar of ggplot2 which will sort the bars according to the numeric variable var.

library(ggplot2)
ggplot(df, aes(var)) +
 geom_bar()

Answer 2

Var data was read as factor because of the percentage symbol. There is no way to read how Numeric using read.table, you need to load the data and then process the string and convert:

dados_ibov <- read.csv2('treinosr/Ibovespa_datah.csv')

dados_ibov$Var <- as.numeric(sub(',', '.', sub('%', '', dados_ibov$Var)))