Hello,
I am wondering if it is possible to compute a rowsums using a list, and also, filter the sum by a condition from a third variable.
I thought in something as
ghj : data frame
list_g=("va1","va2",var3")
ghj$jkj[ ghj$var4==10 ]=apply(ghj[,list_g],1,rowsums)
The syntax before clearly doesn't work.
As you see, I want to sum var1,var2,var3 only if var4==10.
It seems simple, but I can't figure it out.
Can you guide me?
Thanks for your time and consideration
A good example should include data that the R reader will be able to access. So i chose iris dataset that I expect you to have.
library(tidyverse)
list_g <- c("Petal.Length","Sepal.Length")
df_to_rowsum <- select(iris,list_g)
rowsum_result <- rowsum(df_to_rowsum,rep(TRUE,nrow(df_to_rowsum)))
# more tidyverse like
rowsum_result2 <- summarise_all(df_to_rowsum,
sum)
#> rowsum_result
# Petal.Length Sepal.Length
#TRUE 563.7 876.5
#> rowsum_result2
# Petal.Length Sepal.Length
#1 563.7 876.5Thanks, irgrahamuk
But what I desired was a bit more complex.
Suppose, using iris dataset, that I want the rowsum of Sepal.Length, Sepal.Width, and Petal.Length only if Petal.Width is between 0,8 and 1.2.
I applied the next code
zax=datasets::iris
list_g <- c("Petal.Length","Sepal.Width","Petal.Length")
zax$rowsum_result2=ifelse((zax$Petal.Length >=0.8 & zax$Petal.Length <=1.2), rowSums(zax[,list_g]),NA)
table(zax$rowsum_result2)
It seems to work. However, my doubt now is how can you do the same process using the sapply/apply command and not a conditional. Or, in other word, how to include the exception rule in a sapply/apply order.
Again, thanks for your time and interest.
I just don't think apply/sapply etc are the best tools for such jobs
library(tidyverse)
list_g <-c("Sepal.Length","Sepal.Width","Petal.Length")
my_result <- filter(iris,
between(Petal.Width,0.8,1.2)) %>%
summarise_at(list_g,sum)
#> my_result
#Sepal.Length Sepal.Width Petal.Length
# 83 37.7 57.3
this book is highly recommended
I tried your code, but It just collapse the sum vertically.
It should produce a vector with many NA, and only in some, where the condition is met, a value.
It was hard to understand what you wanted but I think I understood now
library(tidyverse)
list_g <-c("Sepal.Length","Sepal.Width","Petal.Length")
my_result <- iris %>%
mutate(rowsum_of_list_g= ifelse(between(Petal.Width,0.8,1.2),
rowSums(.[list_g]),NA))
(show_result <- arrange(my_result,desc(rowsum_of_list_g)) %>% head(n=20))
# Sepal.Length Sepal.Width Petal.Length Petal.Width Species rowsum_of_list_g
# 1 6.1 2.8 4.7 1.2 versicolor 13.6
# 2 5.7 3.0 4.2 1.2 versicolor 12.9
# 3 5.8 2.7 4.1 1.0 versicolor 12.6
# 4 5.5 2.6 4.4 1.2 versicolor 12.5
# 5 5.8 2.7 3.9 1.2 versicolor 12.4
# 6 5.8 2.6 4.0 1.2 versicolor 12.4
# 7 6.0 2.2 4.0 1.0 versicolor 12.2
# 8 5.6 2.5 3.9 1.1 versicolor 12.0
# 9 5.7 2.6 3.5 1.0 versicolor 11.8
# 10 5.5 2.4 3.8 1.1 versicolor 11.7
# 11 5.5 2.4 3.7 1.0 versicolor 11.6
# 12 4.9 2.4 3.3 1.0 versicolor 10.6
# 13 5.0 2.3 3.3 1.0 versicolor 10.6
# 14 5.1 2.5 3.0 1.1 versicolor 10.6
# 15 5.0 2.0 3.5 1.0 versicolor 10.5
# 16 5.1 3.5 1.4 0.2 setosa NA
# 17 4.9 3.0 1.4 0.2 setosa NA
# 18 4.7 3.2 1.3 0.2 setosa NA
# 19 4.6 3.1 1.5 0.2 setosa NA
# 20 5.0 3.6 1.4 0.2 setosa NA
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.