How to create a dataframe with two columns from seperate dataframes in R-CodePudding

I want to create a new dataframe with two columns, each column from a seperate dataframe, and then plot onto the same bar chart

Both columns are called OKS_score.

df1 is preopOKS, this has 645 rows and df2 is postopOKS, this has 643 rows

I have tried left_join

df3 <- preopOKS$OKS_score %>% 
  left_join(postopOKS$OKS_score, copy = TRUE)

I have tried to assign each column and rbind them: however this makes a DF with 2 rows of mulitple values.

x <- preopOKS$OKS_score
y <- postopOKS$OKS_score

z <- rbind(x,y)

str(z)
 num [1:2, 1:645] 36 52 20 34 20 31 33 45 29 51 ...
 - attr(*, "dimnames")=List of 2
  ..$ : chr [1:2] "x" "y"
  ..$ : NULL

I have even tried to pivot_longer to make the df into a format I can plot with:

z %>% 
  pivot_longer(names_to = 'x', values_to = 'y')

As you can tell, I'm very new and can't even get my head around where I'm going wrong, with my methods, or my terribly basic code.

CodePudding user response：

Here's a solution, using the name "OKS" for the original data frame (which is used to generate "preopOKS" and "postopOKS") and assuming it has an "OKS_score" column:

library("dplyr")

# add unique id column to original data frame:
OKS$ID <- 1:nrow(OKS)

# assuming preop data frame has unchanged scores, select just score and ID, 
# but rename score so it's different from postOKS score name:
preopOKS <- OKS[, c("ID", "OKS_score_pre")]

# make postop data frame with whatever function is used to change score:
postopOKS <- data.frame(ID = OKS$ID, OKS_score_post = whatever_function(OKS$OKS_score))

# recombine them
finalOKS <- left_join(preopOKS, postopOKS, by = "ID")

It would be easier, depending on your use case, to just keep the two scores in the same data frame from the start:

library("tidyverse")

OKS <- OKS %>% 
  mutate(OKS_score_pre = OKS_score, 
         OKS_score_post = whatever_function(OKS_score)) %>%
  select(OKS_score_pre, OKS_score_post)

CodePudding user response：

Consider 2 columns:

cola<-runif(20,0,50)

colb<-runif(30,0,50)

Now we want equal column sizes. For that we can attribute the length of the column with maximum size to the other column. Then:

length(cola)<-length(colb)

Now,lets create a new dataframe. For simplicity lets first create a dataframe and name for each column, and combine the two single dataframe columns into one dataframe of two columns:

cola<-data.frame(a=cola)
colb<-data.frame(b=colb)


df<-data.frame(cola,colb)

Then to quickly plot as barplot side by side you can:

par(mfrow=c(1,2))   

barplot(df$a,legend.text = names(df[1]))
barplot(df$b,col="blue",beside=TRUE, legend.text = names(df[2]))

But using ggpplot you can do a better barplot.