R - Using column value in data frame to lookup matrix name and summarize matrix

Question

I am trying to add a new column to an existing data frame which shows the number of one's in a binary matrix. One column in the existing data frame has the the matrix names whose count/sum I am trying to find.

For example,

r <- 10
c <- 10
MatA <- matrix(sample(0:1,r*c, replace=TRUE),r,c)
MatB <- matrix(sample(0:1,r*c, replace=TRUE),r,c)
MatC <- matrix(sample(0:1,r*c, replace=TRUE),r,c)

mat <- c("MatA","MatB","MatC")
size <- c(4,6,10)
df <- data.frame(mat,size)

I need to lookup the mat column values such as MatA, MatB, MatC from df dataframe to match the matrix names MatA, MatB, MatC and return the number of 1's in each of the binary matrix added to the df dataframe in a new column.

I tried using loops, apply functions but am lost at how to use the column values MatA from df$mat as a lookup for matrix name MatA and return sum(MatA==1) to new column in dataframe df .

Answer 1

Use get to reference variable by string

set.seed(7)
df$binary <- lapply(mat, function(x) sum(get(x)))

To use the data.frame column if stored as factors

df$binary <- lapply(levels(df$mat)[df$mat], function(x) sum(get(x)))

To store column as strings and not factors

df <- data.frame(mat, size, stringsAsFactors = FALSE)
df$binary <- lapply(df$mat, function(x) sum(get(x)))

> df
   mat size binary
1 MatA    4     47
2 MatB    6     58
3 MatC   10     54