Multiply rows (with row names) in one data frame w

问题:

I have two data frames:

df1 <- data.frame(Values=c(0.01,0.05), row.names=c("X", "Y"))
df1
  Values
X   0.01
Y   0.05

df2 <-data.frame(c(0,1,1), c(1,0,0), c(1,1,1))
colnames(df2) <- c("X","Y","Z")

df2
  X Y Z
1 0 1 1
2 1 0 1
3 1 0 1

I wish to perform a rowwise operation on df2, where I multiply every column in df2 with its corresponding row in df1 and then perform a summation.

For example, for row 1 of df2, I wish to calculate:

df2 %>% rowwise %>% mutate(newVAL=(df1["X",]*df2[1,"X"])+(df1["Y",]*df2[1,"Y"]))

while excluding columns that don't match (rows in df1) or have NAs.

I have several thousands of rows in df1 and several thousands of rows and columns in df2.

Any help is much appreciated!!

PS. I have implemented this in Perl using hashes and was using the system() call to perform these calculations within an Rmarkdown document. In order to keep it completely reproducible, I am trying to redo it in R. Happy to share the Perl code if necessary.

Thanks.

回答1:

If I understand correctly, it looks like you need sweep.

df3 <- sweep(df2[, rownames(df1)], 2, t(df1), '*')
df3$total <- rowSums(df3)

回答2:

Here's an attempt in base R matching the rows to the columns between the two sets:

rowSums(
  sweep(df2,
        MARGIN=2,
        STATS=df1$Values[match(colnames(df2), rownames(df1))],
        FUN=`*`),
  na.rm=TRUE
)
#[1] 0.05 0.01 0.01

回答3:

We can also use rep to make the lengths same to multiply and then get the rowSums. It will be more efficient to use rep as it is faster

rowSums(df2[rownames(df1)] * rep(df1$Values, each = nrow(df2)))
#[1] 0.05 0.01 0.01

---

Or using the tidyrverse packages

library(dplyr)
library(purrr)
df2 %>% 
     select_(.dots = rownames(df1)) %>% 
     map2(df1$Values, `*`) %>%
     reduce(`+`)
#[1] 0.05 0.01 0.01

Update

If we need it as a column,

df2 %>% 
    select_(.dots = rownames(df1)) %>%
    map2(df1$Values, `*`) %>%
    reduce(`+`) %>%
    mutate(df2, total = .)
#  X Y Z total
#1 0 1 1  0.05
#2 1 0 1  0.01
#3 1 0 1  0.01