CODEWITHSUNDEEP
rggplot2bar-chartdensity-plot
Newbie
Newbie

Reputation: 421

How to get a right "densityplot" in ggplot2?

I have a dataframe in which I have Recall values for 2 different groups of data. I have made a density plot using below code but I am confused about how it looks. For one group (GE in this case) the recall is very high (always 1), in the plot I anticipate a sharp and comparatively higher peak (w.r.t. Epi group) at x-axis=1 while this is not the case in the output graph (see attached picture).

Command I used:

ggplot(epiGE, aes(x=Rec)) + geom_density(aes(group=Class, colour=Class, fill=Class), alpha=0.3) + theme(axis.title.x=element_blank(), axis.title.y=element_blank())

DensityPlot

I have also tried ggplot2 histogram but there as well at x-axis=1 I don't get the right distribution of GE group, as the barplot shows that values in Epi group having Recall=1 are more as compared to GE group.

ggplot(epiGE, aes(x=Rec)) + geom_histogram(aes(group=Class, colour=Class, fill=Class), alpha=0.3) + theme(axis.title.x=element_blank(), axis.title.y=element_blank())

BarPlot

Can somebody please guide me what I am doing wrong here?? Thank you.

Here is my data dput:

> dput(epiGE)
structure(list(Class = structure(c(1L, 2L, 1L, 2L, 1L, 2L, 1L, 
2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 
2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 
2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L), .Label = c("Epi", "GE"), class = "factor"), 
    Rec = c(1, 1, 0.928571428571429, 1, 1, 1, 0.625, 1, 0.935897435897436, 
    1, 0.911764705882353, 1, 0.790697674418605, 1, 0.891891891891892, 
    1, 1, 1, 0.98019801980198, 1, 0.949367088607595, 1, 0.877551020408163, 
    1, 0.944444444444444, 1, 1, 1, 0.955445544554455, 1, 0.5, 
    1, 0.170731707317073, 1, 0.513513513513513, 1, 0, 1, 0.9, 
    1, 0.875, 1, 0.884615384615385, 1, 0.5, 1, 0.9, 1)), .Names = c("Class", 
"Rec"), class = "data.frame", row.names = c(NA, -48L))

Upvotes: 0

Views: 237

Answers (1)

www
www

Reputation: 39174

This is related to the bandwidth of smoothing. You can use the adjust argument to adjust the bandwidth. Here is an example.

ggplot(epiGE, aes(x=Rec)) + 
  geom_density(aes(group=Class, colour=Class, fill=Class), alpha=0.3, adjust = 1/10) + 
  theme(axis.title.x=element_blank(), axis.title.y=element_blank())

Set adjust to 1/10 means using 1/10 of the default bandwidth.

Upvotes: 1

Related Questions