Last active
March 14, 2019 22:27
-
-
Save burchill/6296df75059ef56026a0894eb6f9bcbf to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
library(dplyr) | |
# Copied from magrittr | |
is_pipe <- function (pipe) { | |
identical(pipe, quote(`%>%`)) || identical(pipe, quote(`%T>%`)) || | |
identical(pipe, quote(`%<>%`)) || identical(pipe, quote(`%$%`)) | |
} | |
get_og_lhs <- function(expr) { | |
# While the expression is a call and the first element is a pipe | |
while (is.call(expr) && (is_pipe(expr[[1L]]) || identical(expr[[1L]], quote(`%o%`)))) { | |
expr <- expr[[2L]] | |
} | |
expr | |
} | |
# Takes the end result of everything to the left and the original, | |
# most-left part of the chain, and puts those two values into | |
# the function on the right, in that order. | |
`%o%` <- function(lhs, rhs) { | |
parent <- parent.frame() | |
og_call <- match.call() | |
og_lhs <- get_og_lhs(og_call) | |
rhs_call <- og_call[[3L]] | |
lhs_call <- og_call[[2L]] | |
eval(as.call(c(rhs_call[[1L]], lhs_call, og_lhs, as.list(rhs_call[-1L]))), | |
parent, parent) | |
} | |
# This will just return both values for you to see | |
return_both_values <- function(x,y) { | |
list(old = y, new = x) | |
} | |
# This is an example of a function you could write to 'reset' the attributes | |
set_col_attributes <- function(df, orig_df) { | |
new_cols <- names(df) | |
old_cols <- names(orig_df) | |
for (col in old_cols) { | |
if (col %in% new_cols) { | |
old_attrs <- attributes(orig_df[[col]]) | |
# new_attrs <- attributes(df[[col]]) | |
attributes(df[[col]]) <- old_attrs | |
} | |
} | |
df | |
} | |
# Or you can specify which columns you want | |
keep_these_attributes <- function(df, orig_df, ...) { | |
cols_to_keep <- rlang::ensyms(...) %>% | |
as.character() | |
new_cols <- names(df) | |
old_cols <- names(orig_df) | |
for (col in cols_to_keep) { | |
if (col %in% old_cols) { | |
if (col %in% new_cols) { | |
old_attrs <- attributes(orig_df[[col]]) | |
attributes(df[[col]]) <- old_attrs | |
} | |
} else { | |
warning("'", col, "' not in original data frame") | |
} | |
} | |
df | |
} | |
################################################################################################################ | |
###### EXAMPLES ##################################################################################### | |
################################################################################################################ | |
df <- data.frame(x=1,y=2) | |
class(df) <- c("data.frome", class(df)) | |
attributes(df$x) <- list(class="myclass", something="my.attribute") | |
df1 <- df %>% | |
mutate(z=3) %>% | |
group_by(y) %>% | |
mutate(x=x/2) %>% | |
ungroup() | |
attributes(df1$x) | |
df2 <- df %>% | |
mutate(z=3) %>% | |
group_by(y) %>% | |
mutate(x=x/2) %>% | |
ungroup() %o% | |
set_col_attributes() | |
attributes(df2$x) | |
df3 <- df %>% | |
mutate(z=3) %>% | |
group_by(z) %>% | |
mutate(x=x/2, y=y+4) %>% | |
ungroup() %o% | |
keep_these_attributes(x) | |
attributes(df3$x) | |
attributes(df3$y) | |
df %>% | |
mutate(z=3) %>% | |
group_by(y) %>% | |
summarise(n=n()) %o% | |
return_both_values() | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment