CODEWITHSUNDEEP
rplotly
Pedro Elizalde Acevedo
Pedro Elizalde Acevedo

Reputation: 69

Change each histogram color in chart

I have this histogram separated in five categories depending the age. The problem is that I cannot change the color depending the category. I tried to use the marker function with an array filled with the colors I want, but it didn't work as expected:

Bugged Histogram

As you can see, the colors are all bugged.

This is what I tried:

less20 <- subset(dataset, dataset$EDAD <20)
between20n40 <- subset(dataset, dataset$EDAD >=20 & dataset$EDAD <40)
between40n60 <- subset(dataset, dataset$EDAD >=40 & dataset$EDAD <60)
between60n80 <- subset(dataset, dataset$EDAD >=60 & dataset$EDAD <80)
more80 <- subset(dataset, dataset$EDAD >=80)

plot_ly(alpha = 0.7, orientation = 'h', marker = list(color = c('rgba(31,119,180,1)','rgba(105,122,125,1)','rgba(183,124,67,1)', 'rgba(243,127,23,1)','rgba(255,127,14,1)'))) %>%
  add_histogram(y = more80$EDAD, name = "More than 80") %>%
  add_histogram(y = between60n80$EDAD, name = "Between 60 and 79") %>%
  add_histogram(y = between40n60$EDAD, name = "Between 40 and 59") %>%
  add_histogram(y = between20n40$EDAD, name = "Between 20 and 39") %>%
  add_histogram(y = less20$EDAD, name = "Less than 20") %>%
layout(barmode = "group", title = "",orientation="h")

The correct color order is the next one: Correct color order but not the colors I'm looking for However, I want to change those colors.

Any recomendations? Thanks in advance :)

Upvotes: 0

Views: 528

Answers (2)

Pedro Elizalde Acevedo
Pedro Elizalde Acevedo

Reputation: 69

I just had to add the marker function inside add_histogram. That way, I only change the color of each histogram added.

   plot_ly(alpha = 0.7, orientation = 'h') %>%
add_histogram(y = more80$EDAD, name = "More than 80", marker = list(color ='rgba(31,119,180,1)')) %>%
  add_histogram(y = between60n80$EDAD, name = "Between 60 and 79", marker = list(color ='rgba(105,122,125,1)')) %>%

Thank you for your answers!

Upvotes: 1

rafagarci
rafagarci

Reputation: 97

I think it might be easier if you put the values along with their corresponding ranges inside a dataframe and color the plot using these values and ranges. This is my solution to this using ggplot2. You can define Values with your original dataset for the histogram and should obtain a similar result to yours. This solution uses a 1000 normally distributed sample with SD = 30 and MEAN = 70 in order to produce the plot.

# Import ggplot2
library("ggplot2")

# Obtain sample values for histogram
set.seed(1234)
Values = rnorm(n = 1000, mean = 70, sd = 30)
Range = c()

# Get ranges for each value in data
for(i in 1:length(Values)){
  if(Values[i] >= 80){
    Range[i] = "More than 80"
  } else if (Values[i] < 80 & Values[i] >= 60){
    Range[i] = "Between 60 and 79"
  } else if (Values[i] < 60 & Values[i] >= 40){
    Range[i] = "Between 40 and 59"
  } else if (Values[i] < 40 & Values[i] >= 20){
    Range[i] = "Between 20 and 39"
  } else {
    Range[i] = "Less than 20"
  }
}


# Put all data inside a data frame
plot_dat = data.frame(Values, Range)

# Order plot labels
plot_dat$Range <- factor(plot_dat$Range, levels = c("More than 80", "Between 60 and 79", "Between 40 and 59", "Between 20 and 39", "Less than 20"))

# Produce plot
ggplot(plot_dat, aes(x=Values, fill=Range)) + geom_histogram(binwidth = 5) +  coord_flip() + ggtitle("Sample Histogram")

Output

Upvotes: 1

Related Questions