T-T T-T - 10 days ago 5
R Question

Repeating calculation based on conditions

What I am trying to do is pretty simple. However, I am new to R and have not learned much about loops and functions and am not sure what is the most efficient way to get the results. Basically, I want to count the number of rows that meet my conditions and do a division. Here is an example:

df1 <- data.frame(
Main = c(0.0089, -0.050667, -0.030379, 0.066484, 0.006439, -0.026076),
B = c(NA, 0.0345, -0.0683, -0.052774, 0.014661, -0.040537),
C = c(0.0181, 0, -0.056197, 0.040794, 0.03516, -0.022662),
D = c(-0.0127, -0.025995, -0.04293, 0.057816, 0.033458, -0.058382)
)
df1
# Main B C D
# 1 0.008900 NA 0.018100 -0.012700
# 2 -0.050667 0.034500 0.000000 -0.025995
# 3 -0.030379 -0.068300 -0.056197 -0.042930
# 4 0.066484 -0.052774 0.040794 0.057816
# 5 0.006439 0.014661 0.035160 0.033458
# 6 -0.026076 -0.040537 -0.022662 -0.058382


My criteria for the numerator is to count the number of
B/C/D
that is
>0
when
Main
is
>0
; For denominator, count the number of
B/C/D
that is
!= 0
when
Main
is
!= 0
. I can use
length(which(df1$Main >0 & df1$B>0)) / length(which(df1$Main !=0 & df1$B !=0))
to get the ratios for each of the column individually. But my data set has many more columns, and I am wondering if there is a way to get those ratio all at once so that my result will be like:

# B C D
# 1 0.2 0.6 0.3

Answer

Use apply:

 apply(df1[,-1], 2, function(x) length(which(df1$Main >0 & x>0)) / length(which(df1$Main !=0 & x !=0)))