Calculate and merge vector output with scalar input using purrr or plyr in R

Question

I have the following code calculate output from scalar input and merge the results into a data frame together with the input. I tried to do the same thing with purrr and plyr, but got stuck. Can anyone provide some equivalent code?

library(data.table)
library(magrittr)

sigma = rep(2,5)

sim_ar1 = function(b1,sigma) b1 + sigma

data.table(b1 = seq(0.9,1,0.03)) %>% 
  .[,
    data.frame(sigma, sim1 = sim_ar1(b1,sigma)), 
    by = b1]

Answer 1

Here is a solution using tidyverse. The key is to use expand.grid to create all the combination between b1 and sigma, and then use map2 to apply the sim_ar1 function to create the sim1 column. dt should look like the same as your example output from data.table.

library(tidyverse)

sigma <- rep(2,5)
b1 <- seq(0.9,1,0.03)
sim_ar1 <- function(b1,sigma) b1 + sigma

dt <- expand.grid(b1 = b1, sigma = sigma) %>%
  arrange(b1) %>%
  mutate(sim1 = map2(b1, sigma, sim_ar1))

dt
     b1 sigma sim1
1  0.90     2  2.9
2  0.90     2  2.9
3  0.90     2  2.9
4  0.90     2  2.9
5  0.90     2  2.9
6  0.93     2 2.93
7  0.93     2 2.93
8  0.93     2 2.93
9  0.93     2 2.93
10 0.93     2 2.93
11 0.96     2 2.96
12 0.96     2 2.96
13 0.96     2 2.96
14 0.96     2 2.96
15 0.96     2 2.96
16 0.99     2 2.99
17 0.99     2 2.99
18 0.99     2 2.99
19 0.99     2 2.99
20 0.99     2 2.99

Update

In fact, since sim_ar1 is vectorized, there is no need to use map2 from purrr. dplyr alone is sufficient.

library(dplyr)

dt <- expand.grid(data_frame(b1 = b1, sigma = 2)) %>%
  arrange(b1) %>%
  mutate(sim1 = sim_ar1(b1, sigma))

Answer 2

I also develop an answer using map with bind_rows and would like to share here

library(data.table)
library(magrittr)

sigma = rep(2,5)

sim_ar1 = function(b1,sigma) b1 + sigma

map(b1, ~data.frame(b1 = .x, results = sim_ar1(.x,sigma))) %>%  bind_rows()

As is suggested by @ycw, using map_df could make the come more concise and the last line would be:

map_df(b1, ~data.frame(b1 = .x, results = sim_ar1(.x,sigma)))