Notes for tidy evaluation based on strings as sources of column names

tidy_eval_notes.R

# Tidyeval notes
library(dplyr)
library(rlang)
library(tasic2016data)

# one character element:
test <- "primary_type_id"

expr_test <- parse_expr(test)

tasic_2016_anno %>%
  group_by(!!expr_test) %>%
  summarise(x = n())
# or
tasic_2016_anno %>%
  group_by(!!parse_expr(test)) %>%
  summarise(x = n())

# one or more character elements
test <- c("primary_type_id","secondary_type_id")

expr_test <- lapply(test, parse_expr)

tasic_2016_anno %>%
  group_by(!!!expr_test) %>%
  summarise(x = n())

# a single entry from multiple elements
test <- c("primary_type_id","secondary_type_id")

expr_test <- lapply(test, parse_expr)

tasic_2016_anno %>%
  group_by(!!expr_test[[1]]) %>%
  summarise(x = n())

# Filtering:
test <- "primary_type_id"
test_targets <- c(1:5)

tasic_2016_anno %>%
  filter(!!parse_expr(test) %in% test_targets) %>%
  group_by(!!parse_expr(test)) %>%
  summarise(x = n())
# or
expr_test <- parse_expr(test)

tasic_2016_anno %>%
  filter(!!expr_test %in% test_targets) %>%
  group_by(!!expr_test) %>%
  summarise(x = n())

Note: the tasic2016data package can be installed with:

devtools::install_github("AllenInstitute/tasic2016data")