drop_lvls2: Drop factor levels and filter the data the same time

In emilelatour/lamisc: Latour's Little Helpers

Drop factor levels and filter the data the same time

Description

When you filter data to remove rows that match a level in a factor, the factor levels aren't removed. This function will filter the data and drop the factor levels.

Usage

drop_lvls2(data, var, lvls_to_drop = NULL)

Arguments

data	A data frame or tibble. Or a survey object (class = tbl_svy, survey.design2, or survey.design)
var	A factor (or character vector).
lvls_to_drop	Character vector of factors to drop

Value

An object of the same type as data.

Examples

library(dplyr)
library(tibble)
library(palmerpenguins)
library(srvyr)
library(survey)

# With a data frame
df <- tibble::tibble(letters = letters[1:5],
                     numbers = seq(1:5))

f_lvls <- letters[1:6]

df <- df %>%
  mutate(letters = factor(letters,
                          levels = f_lvls))

levels(df$letters)

df %>%
  drop_lvls2(data = .,
            var = letters,
            lvls = "a") %>%
  dplyr::count(letters,
               .drop = FALSE)

df %>%
  drop_lvls2(data = .,
            var = letters,
            lvls = c("a", "b", "g")) %>%
  dplyr::count(letters,
               .drop = FALSE)


# With a tibble
penguins %>%
  dplyr::count(species)

penguins %>%
  drop_lvls2(data = .,
            var = species,
            lvls = c("Chinstrap")) %>%
  dplyr::count(species)

penguins %>%
  drop_lvls2(data = .,
            var = species,
            lvls = c("Chinstrap",
                     "Gentoo")) %>%
  dplyr::count(species)


# With a survey using srvyr package
data(api)

srs_design_srvyr <- apisrs %>% as_survey_design(ids = 1, fpc = fpc)
srs_design_survey <- svydesign(ids = ~1, fpc = ~fpc, data = apisrs)

srs_design_srvyr %>%
  group_by(stype) %>%
  summarize(proportion = survey_mean(),
            total = survey_total())

srs_design_srvyr %>%
  drop_lvls2(data = .,
            var = stype,
            lvls_to_drop = "H") %>%
  group_by(stype) %>%
  summarize(proportion = survey_mean(),
            total = survey_total())

srs_design_srvyr %>%
  drop_lvls2(data = .,
            var = stype,
            lvls_to_drop = c("H", "M")) %>%
  group_by(stype) %>%
  summarize(proportion = survey_mean(),
            total = survey_total())

# With a survey using survey package
svymean(~stype, srs_design_survey)
svytotal(~stype, srs_design_survey)

srs_design_survey2 <- srs_design_survey %>%
  drop_lvls2(data = .,
            var = stype,
            lvls_to_drop = c("H"))

svymean(~stype, srs_design_survey2)
svytotal(~stype, srs_design_survey2)

emilelatour/lamisc documentation built on April 9, 2024, 10:33 a.m.