Merging different data frames with different columns in R

Question

I tried with the following code:

library(ggplot2)
library(dplyr)
library(tidyr)

other <- read.table(text='Phase Year    V14 V15
Other   2017    0.016488414 0.027183601
Other   2016    0.016937669 0.016937669
Other   2016    0.020214521 0.010313531
Other   2016    0.025205761 0.099279835
Other   2016    0.014341085 0.037596899
Other   2016    0.01622807  0.04254386', header=TRUE)


code <- read.table(text='Phase   Year    V5  V8  V19 V21
Code    2017    0.016488414 0.053921569 0.01114082  0.027183601
Code    2016    0.033197832 0.016937669 0.016937669 0.016937669', header=TRUE)

test <- read.table(text='Phase   Year    V6  V11 V20 V22
Test    2017    0.032531194 0.027183601 0.016488414 0.305258467
Test    2016    0.106368564 0.025067751 0.016937669 0.025067751', header=TRUE)

# tidy data
code_df <- code %>%
          gather(key, value, V5, V8, V19, V21)
test_df <- test %>%
          gather(key, value, V6, V11, V20, V22)
other_df <- other %>%
          gather(key, value, V14, V15)


newd <- merge(other_df, code_df, test_df, all=TRUE)
par(mfrow = c(1, 2))
ggplot(data=newd, aes(x=Year)) +
  geom_density(aes(group=Phase, color=Phase, fill=Phase, showguide=FALSE)) +
  ggtitle("Test")+
  xlab("Year")+
  ylab("Probability")+
  facet_wrap(~Phase, ncol=1)

My results does not look promising, I would like to plot my data set without loosing any of them after the merging, because I want each data frame in a separated graph, but all the same in the same frame. Does any one have an idea about what what is wrong in my code?

Please make your question reproducible by posting your dataset with `dput(df)` — acylam, Sep 17 '17 at 04:04
See this to learn how to make good R reproducible example: https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example — www, Sep 17 '17 at 04:08

Kevin Arseneau · Accepted Answer · 2017-09-17T04:54:38.387

1

I think you are just trying to bind the columns together based on both Phase and Year. The dplyr approach would be.

library(dplyr)

df <- df1 %>%
  bind_cols(df2, by = c("Phase", "Year")) %>%
  bind_cols(df3, by = c("Phase", "Year"))

Edit

I now see with your gather that you are building a long format, so we'll switch to bind_rows to bring together.

newd <- bind_rows(other_df, code_df, test_df)

Your plot will now generate after changing that line.

edited Sep 17 '17 at 04:54

answered Sep 17 '17 at 04:20

Kevin Arseneau

6,186
1
21
40

It gives me an error `Error in list_or_dots(...) : object 'Phase' not found ` – Sultan Sep 17 '17 at 04:26
@Sultan, your question is not reproducible, I have no way of knowing what contributed to your error. – Kevin Arseneau Sep 17 '17 at 04:27
How to make it reproducible, I really got confused – Sultan Sep 17 '17 at 04:29
@Sultan, there was a link posted above by ycw that details how to do it right. – Kevin Arseneau Sep 17 '17 at 04:32
@Sultan, I have edited with quoted variables after capturing your data. I will add to my answer so that you can see how it is done. – Kevin Arseneau Sep 17 '17 at 04:41
I have updated the question and now it is reproducible, it think. – Sultan Sep 17 '17 at 04:42
@Sultan, that is much better. You should always make it easy for other people to help you. – Kevin Arseneau Sep 17 '17 at 04:43

Merging different data frames with different columns in R

1 Answers1

Edit