dummy <- data.frame(Q1 = c(0, 1, 0, 1),
Q2 = c(1, 1, 0, 1),
Q3 = c(0, 1, 1, 0))
df_dummy <- data.frame(Question = c("Q1", "Q2", "Q3"),
X1 = c(2/4, 3/4, 2/4),
X0 = c(2/4, 1/4, 2/4))
> dummy
Q1 Q2 Q3
1 0 1 0
2 1 1 1
3 0 0 1
4 1 1 0
> df_dummy
Question X1 X0
1 Q1 0.50 0.50
2 Q2 0.75 0.25
3 Q3 0.50 0.50
I have some data (dummy) where I have binary responses to Q1, Q2, and Q3. I want to summarize my data in the format as shown in df_dummy, where for each question, column X1 tells me the proportion of people that answered 1 to Q1, and column X0 tells me the proportion of people that answered 0 to Q0. I tried prop.table but that didn't return the desired result.
Another way is counting the proportion of 1s and then deducing from that the proportion of 0s:
X1 <- colSums(dummy==1)/nrow(dummy)
df_dummy <- data.frame(X1, X0=1-X1)
df_dummy
# X1 X0
#Q1 0.50 0.50
#Q2 0.75 0.25
#Q3 0.50 0.50
NB, inspired from @akrun's idea of ColMeans: You can also use colMeans instead of dividing colSumsby the number of row to define X1:
X1 <- colMeans(dummy==1)
df_dummy <- data.frame(X1, X0=1-X1)
df_dummy
# X1 X0
#Q1 0.50 0.50
#Q2 0.75 0.25
#Q3 0.50 0.50