Chargaff
Chargaff

Reputation: 1572

plotting melted data.frame with ggplot, does aes(x) need to be a factor?

I don't understand why ggplot can't plot this data.frame :

sample <- dput(melted)
structure(list(alleles = c(98, 100, 102, 106, 124, 126, 128, 
132, 134, 142, 144, 145, 146, 147, 148, 149, 151, 152, 153, 156, 
158, 159, 165, 167, 169, 171, 173, 175, 177, 181, 184, 187, 193, 
194, 196, 197, 200, 228, 233, 234, 238, 240, 241, 242, 243, 244, 
245, 246, 247, 249, 251, 253, 363, 364, 365, 367, 371, 377, 380, 
384, 391, 98, 100, 102, 106, 124, 126, 128, 132, 134, 142, 144, 
145, 146, 147, 148, 149, 151, 152, 153, 156, 158, 159, 165, 167, 
169, 171, 173, 175, 177, 181, 184, 187, 193, 194, 196, 197, 200, 
228, 233, 234, 238, 240, 241, 242, 243, 244, 245, 246, 247, 249, 
251, 253, 363, 364, 365, 367, 371, 377, 380, 384, 391), variable = structure(c(1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L), .Label = c("Parents", "Nymphes"
), class = "factor"), value = c(11, 17, 13, 1, 1, 18, 5, 8, 10, 
6, 9, 1, 25, 17, 2, 1, 1, 2, 0, 2, 0, 2, 1, 1, 8, 9, 6, 10, 35, 
1, 3, 4, 20, 3, 2, 39, 2, 0, 0, 1, 8, 25, 2, 27, 6, 8, 0, 3, 
1, 8, 1, 2, 14, 17, 2, 5, 1, 2, 1, 2, 0, 12, 49, 33, 0, 4, 35, 
6, 7, 14, 7, 25, 0, 44, 26, 3, 0, 6, 0, 9, 3, 1, 0, 1, 4, 15, 
22, 8, 24, 69, 0, 2, 5, 35, 2, 18, 92, 0, 2, 6, 0, 22, 44, 6, 
56, 13, 12, 1, 6, 2, 21, 0, 3, 12, 9, 5, 3, 0, 1, 1, 0, 1)), .Names = c("alleles", 
"variable", "value"), row.names = c(NA, -122L), class = "data.frame")

ggplot command :

ggplot(sample,aes(x=alleles,y=value)) + geom_bar(aes(fill=variable),position="dodge") + theme(axis.text.x=element_text(angle = -75, hjust = 0))

error :

stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this. Error in pmin(y, 0) : object 'y' not found

If first column of data.frame (alleles) is a factor, then the plot comes out, but then I have problems with the ordering of data. I need column alleles to be in order, as in the above data.frame. There is obviously something that escapes me...

Upvotes: 1

Views: 1424

Answers (2)

cbeleites
cbeleites

Reputation: 14093

If this is what you are looking for:

enter image description here

you need to set stat = "identity" in geom_bar

ggplot(sample,aes(x=alleles + (as.numeric (variable)/2 - .75), y=value)) + 
    geom_bar(aes(fill=variable),position="identity", stat = "identity") + 
    theme(axis.text.x=element_text(angle = -75, hjust = 0))

geom_bar defaults to counting (i.e. makeing a histogram), but if I understood you correctly, the counts (or values) are already calculated.

Upvotes: 1

Arun
Arun

Reputation: 118889

I think you're using geom_bar wrongly. If you already have the frequency (or length of the bar) calculated, you should use the weight = aesthetic as follows:

sample$alleles <- factor(sample$alleles)
ggplot(sample, aes(x = alleles)) + 
    geom_bar(aes(weight = value, fill = variable), position = "dodge") + 
    theme(axis.text.x=element_text(angle = -75, hjust = 0))

If by factoring it doesn't order the plot in the order you desire, then you can order your factor in this manner (if you want the numerical order preserved):

sample$alleles <- factor(sample$alleles, 
                levels=sample$alleles[!duplicated(sample$alleles)], ordered = T)

Now, plotting should result in the numerical order. If you want to change the order to something else, set the levels = appropriately.

After reordering and plotting, this is what I get:

ggplot2_weight

Upvotes: 1

Related Questions