Generate a boxplot for grouped data using ggplot2

Question

Sorry if this is so basic.

I am trying to use ggplot2 to draw a boxplot for my data. However, I am having troubles of melting the data into the correct format.

Here is how my data looks like:

head(Final_RMSE_MC)

  N_reg N_var RMSE_MC_1 RMSE_MC_2 RMSE_MC_3 RMSE_MC_4 RMSE_MC_5 RMSE_MC_6 RMSE_MC_7 RMSE_MC_8 RMSE_MC_9 RMSE_MC_10 RMSE_MC_11 RMSE_MC_12 RMSE_MC_13
1     6     5 0.5016800 0.5898132 0.5482860 0.4585713 0.4830320 0.4376286 0.4626646 0.5753290 0.4600453  0.4625784  0.4135086  0.5356082  0.4262005
2    10     5 0.4928764 0.4426350 0.4634775 0.4049509 0.5192989 0.4420706 0.3912822 0.4808609 0.4190173  0.4828170  0.4123871  0.4394507  0.4100748
3     4     4 0.4890946 0.4503532 0.5930480 0.5608510 0.4232696 0.5392966 0.4134308 0.5950408 0.5425955  0.5209573  0.6669176  0.4819051  0.4926042
4     5     4 0.5229090 0.5076377 0.5254299 0.4455789 0.4816532 0.5468765 0.4474718 0.4467224 0.4280381  0.6339686  0.3921858  0.5335065  0.4548194
5     9     4 0.4138625 0.4782089 0.4522069 0.4534526 0.4175361 0.4685324 0.3908619 0.4877251 0.4509520  0.4410600  0.4685804  0.4660575  0.4775753
6     3     3 0.5135749 0.6280533 0.5841148 0.5051640 0.5279784 0.5981735 0.4638461 0.4664253 0.4568787  0.4150206  0.5780827  0.5474891  0.4232878

I would like to melt the data based on N_var for all the columns except N_reg and N_var. So I tried

dfm <- melt(Final_RMSE_MC, id.vars = "N_reg")

and I got

  N_reg variable value
1     6    N_var     5
2    10    N_var     5
3     4    N_var     4
4     5    N_var     4
5     9    N_var     4
6     3    N_var     3

which does not seem right.

My next step is to use

ggplot(dfm, aes(x = variable, y = value)) + geom_boxplot()

Thanks for any suggestions!

Here's how to make Tufte's variant: stackoverflow.com/questions/6973394/… — Ari B. Friedman
– Ari B. Friedman, Commented May 2, 2013 at 8:34

Didzis Elferts · Accepted Answer · 2013-05-02 08:18:34Z

1

You should use both variables N_reg and N_var as id.vars as they are the same for all other variables in one row.

dfm <- melt(Final_RMSE_MC, id.vars = c("N_reg","N_var"))

head(dfm)
  N_reg N_var  variable     value
1     6     5 RMSE_MC_1 0.5016800
2    10     5 RMSE_MC_1 0.4928764
3     4     4 RMSE_MC_1 0.4890946
4     5     4 RMSE_MC_1 0.5229090
5     9     4 RMSE_MC_1 0.4138625
6     3     3 RMSE_MC_1 0.5135749

ggplot(dfm, aes(x = variable, y = value)) + geom_boxplot()

edited May 2, 2013 at 8:18

answered May 2, 2013 at 8:08

Didzis Elferts

99.2k17 gold badges275 silver badges207 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

TTT Over a year ago

Thanks. But actually N_reg and N_var are different

Didzis Elferts Over a year ago

I ment that they are the same in one row.

TTT Over a year ago

I see. should I use ggplot(dfm, aes(x = N_reg, y = value)) + geom_boxplot() if I want to build boxplot for values with the same N_reg?

TTT Over a year ago

It is factor(N_reg). I found it. Thanks so much!

Collectives™ on Stack Overflow

Generate a boxplot for grouped data using ggplot2

1 Answer 1

4 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related