multiple-rollers.R

suppressPackageStartupMessages(library(tibbletime))
suppressPackageStartupMessages(library(tidyverse))

data(FB)

# Create the column names
col_names <- map_chr(2:10, ~paste0("adjusted_", .x))

# Creating the rolling functions and assign them names
rollers <- map(2:10, ~rollify(mean, window = .x)) %>%
  set_names(nm = col_names)

# Use invoke map, it takes a list of functions and calls each one with the 
# supplied args
FB_with_multiroll <- bind_cols(FB, invoke_map(rollers, x = FB$adjusted))

FB_with_multiroll %>% select(starts_with("adjusted"))
#> # A tibble: 1,008 x 10
#>    adjusted adjusted_2 adjusted_3 adjusted_4 adjusted_5 adjusted_6
#>       <dbl>      <dbl>      <dbl>      <dbl>      <dbl>      <dbl>
#>  1    28.00         NA         NA         NA         NA         NA
#>  2    27.77     27.885         NA         NA         NA         NA
#>  3    28.76     28.265   28.17667         NA         NA         NA
#>  4    29.42     29.090   28.65000    28.4875         NA         NA
#>  5    29.06     29.240   29.08000    28.7525     28.602         NA
#>  6    30.59     29.825   29.69000    29.4575     29.120   28.93333
#>  7    31.30     30.945   30.31667    30.0925     29.826   29.48333
#>  8    31.72     31.510   31.20333    30.6675     30.418   30.14167
#>  9    30.95     31.335   31.32333    31.1400     30.724   30.50667
#> 10    30.10     30.525   30.92333    31.0175     30.932   30.62000
#> # ... with 998 more rows, and 4 more variables: adjusted_7 <dbl>,
#> #   adjusted_8 <dbl>, adjusted_9 <dbl>, adjusted_10 <dbl>

This worked great for me but now I'd like to do this on multiple groups of FB$adjusted. For instance, if there were multiple stock symbols in this example across the same date range, I'd like to have the same multiple rolling means calculated for each group.

I tried the equivalent of:
FB_with_multiroll <- FB%>%
group_by(symbol) %>%
bind_cols(FB, invoke_map(rollers, x = FB$adjusted))

But that didn't work.

Thanks