CODEWITHSUNDEEP
rggplot2
K_abraham
K_abraham

Reputation: 1

Change manual color on my geom_plot

I have an issues and have searching and searching. I'm really really new to this R Code. I have 2 different values from my gender array 1=Male and 2=Female.

I can't figure out how to change the scale from gradient to 1 and 2 color, and how I can make different color to the female and male. I close, but can't do the last job :(

ggplot(smokingdata, aes(x=ages, y=consume, col=gender)) +
geom_point() + ylim(0, 80)

Hop there is someone how can help me with that.

Upvotes: 0

Views: 367

Answers (2)

W. Murphy
W. Murphy

Reputation: 1141

The real issue here is that your data is incorrectly specified. gender is a discrete variable that is stored as a numeric value, so ggplot2 is treating it as a continuous variable. Simply convert this variable to a factor and you will get a discrete color scale.

ggplot(smokingdata, aes(x=ages, y=consume, col=as.factor(gender))) +
geom_point() + ylim(0, 80)

However, this won't generate a very useful legend. Modifying your data frame to represent the variable accurately would yield the best result:

smokingdata$gender <- factor(smokingdata$gender, levels = c(1,2), labels = c("Male", "Female"))
ggplot(smokingdata, aes(x=ages, y=consume, col=gender)) +
geom_point() + ylim(0, 80)

Upvotes: 1

camille
camille

Reputation: 16832

You can give a vector, either named or unnamed, to scale_color_manual. So if your column for gender is encoded as 1 & 2, you could use scale_color_manual(values = c("1" = "purple", "2" = "orange")). Note that if you're using numbers for gender, those need to be turned into characters to work as names in the named vector.

You could also change the gender column to a character vector instead of a numeric one; that way, ggplot will treat it as a discrete variable, rather than a continuous one.

Upvotes: 0

Related Questions