Awesome

unnestIfElse

RStudio Addin to quickly and roughly convert nested ifelse() statements to dplyr::case_when().

Demonstration of add-in

Contents

Installation and requirements
Before and after
Motivation
Features
Limitations
Examples
TODO
License

Installation and requirements

Install

devtools::install_github("erictleung/unnestIfElse")

Requirements

dplyr
magrittr
stringr
rstudioapi

Before and after

Before:

ifelse(statement, TRUE, FALSE)

After:

case_when(
  statement ~ TRUE
  TRUE ~ FALSE
)

Motivation

Inspired by:

<blockquote class="twitter-tweet"><p lang="en" dir="ltr"><a href="https://twitter.com/hashtag/rstats?src=hash&ref_src=twsrc%5Etfw">#rstats</a> does someone have a magic function to turn nested ifelse statements into dplyr::casewhen?</p>— Gordon Shotwell (@gshotwell) <a href="https://twitter.com/gshotwell/status/1187783316038148096?ref_src=twsrc%5Etfw">October 25, 2019</a></blockquote>

Features

Changes highlighted text in place
Highlighted text should be complete ifelse() expression

Limitations

Only works for highlighted text, cannot convert an entire script
Highlighted text has to exclusively include complete ifelse() statement
Lacking any error handling
Doesn't handle the case if the last "no" value is some function call with parentheses (e.g., ifelse(1 != 1, -4, abs(-4)))
Limited test coverage for use
Code formatting will fail with any text with commas (see below)

# Addin will fail with this example
# Source:
# https://github.com/Sean-Hughes/Relational_Complexity_Derivation/blob/master/Analyses/independent%20t%20test%20-%20IAT%20D1.R
nhst <- ifelse(t_test_p < 0.05, 
               paste("A Welch's independent t test demonstrated significant differences of ", d_interpretation, " effect size in ", DV_name, " between ", sep = ""),
               paste("A Welch's independent t test demonstrated non-significant differences of ", d_interpretation, " effect size between ", sep = ""))

# Another example that will fail
# Source: https://github.com/wmlutz/r-lesson/blob/master/hold_temp.R
aw$cert_Admin <- ifelse("Admin" %in% raw[,"Certifications"], 1,0)

Examples

I did a quick advanced GitHub search of ifelse language:R to find these examples. I've tried to pick a mix of short to long examples. I also wanted to make sure this function works on code in the "wild".

Simple

Source: https://github.com/Anis-Bensaid/SBT13/blob/master/Script%20R%20Final.R

Before:

data$a1 <- ifelse(bd1$A1=="Pas du tout", 0, ifelse(bd1$A1=="Un peu", 1,ifelse(bd1$A1== "Beaucoup", 2,ifelse(bd1$A1 == "Enormément", 3,NA))))

After:

data$a1 <- case_when(
  bd1$A1=="Pas du tout" ~  0,
   bd1$A1=="Un peu" ~  1,
  bd1$A1== "Beaucoup" ~  2,
  bd1$A1 == "Enormément" ~  3,
  TRUE ~ NA
)

Mix of numbers

Source: https://github.com/ajpelu/qpyr_resilience/blob/master/R/getComposite.R

Before:

getComposite_nonleap <- function(m,d){
  # m = month
  # d = day
  ifelse(m == 1 & d < 17, 1,
  ifelse(m == 1 & d >= 17, 2,
  ifelse(m == 2 & d < 2, 2,
  ifelse(m == 2 & d < 18, 3,
  ifelse(m == 2 & d >= 18, 4,
  ifelse(m == 3 & d < 6, 4,
  ifelse(m == 3 & d < 22, 5,
  ifelse(m == 3 & d >= 22, 6,
  ifelse(m == 4 & d < 7, 6,
  ifelse(m == 4 & d < 23, 7,
  ifelse(m == 4 & d >= 23, 8,
  ifelse(m == 5 & d < 9, 8,
  ifelse(m == 5 & d < 25, 9,
  ifelse(m == 5 & d >= 25, 10,
  ifelse(m == 6 & d < 10, 10,
  ifelse(m == 6 & d < 26, 11,
  ifelse(m == 6 & d >= 26, 12,
  ifelse(m == 7 & d < 12, 12,
  ifelse(m == 7 & d < 28, 13,
  ifelse(m == 7 & d >= 28, 14,
  ifelse(m == 8 & d < 13, 14,
  ifelse(m == 8 & d < 29, 15,
  ifelse(m == 8 & d >= 29, 16,
  ifelse(m == 9 & d < 14, 16,
  ifelse(m == 9 & d < 30, 17,
  ifelse(m == 9 & d >= 30, 18,
  ifelse(m == 10 & d < 16, 18,
  ifelse(m == 10 & d >= 16, 19,
  ifelse(m == 11 & d < 17, 20,
  ifelse(m == 11 & d >= 17, 21,
  ifelse(m == 12 & d < 3, 21,
  ifelse(m == 12 & d < 19, 22, 23))))))))))))))))))))))))))))))))
}

After:

getComposite_nonleap <- function(m,d){
  # m = month 
  # d = day 
  case_when(
  m == 1 & d < 17 ~ 1,
  m == 1 & d >= 17 ~ 2,
  m == 2 & d < 2 ~ 2,
  m == 2 & d < 18 ~ 3,
  m == 2 & d >= 18 ~ 4,
  m == 3 & d < 6 ~ 4,
  m == 3 & d < 22 ~ 5,
  m == 3 & d >= 22 ~ 6,
  m == 4 & d < 7 ~ 6,
  m == 4 & d < 23 ~ 7,
  m == 4 & d >= 23 ~ 8,
  m == 5 & d < 9 ~ 8,
  m == 5 & d < 25 ~ 9,
  m == 5 & d >= 25 ~ 10,
  m == 6 & d < 10 ~ 10,
  m == 6 & d < 26 ~ 11,
  m == 6 & d >= 26 ~ 12,
  m == 7 & d < 12 ~ 12,
  m == 7 & d < 28 ~ 13,
  m == 7 & d >= 28 ~ 14,
  m == 8 & d < 13 ~ 14,
  m == 8 & d < 29 ~ 15,
  m == 8 & d >= 29 ~ 16,
  m == 9 & d < 14 ~ 16,
  m == 9 & d < 30 ~ 17,
  m == 9 & d >= 30 ~ 18,
  m == 10 & d < 16 ~ 18,
  m == 10 & d >= 16 ~ 19,
  m == 11 & d < 17 ~ 20,
  m == 11 & d >= 17 ~ 21,
  m == 12 & d < 3 ~ 21,
  m == 12 & d < 19 ~ 22,
  TRUE ~ 23
)
}

Text and numbers

Source: https://github.com/WWF-ConsEvidence/ACES/blob/master/Mozambique/Harmonized_Demographic_Data.R

Before:

Community2008 <- ifelse(dat2008$Community_code == 241601, "Fuzi", 
             ifelse(dat2008$Community_code == 41602, "Moebase",
             ifelse(dat2008$Community_code==41603, "Txotxo",
             ifelse(dat2008$Community_code ==41604, "Monudo/Murudo", 
             ifelse(dat2008$Community_code ==41605, "Namige", 
             ifelse(dat2008$Community_code ==41606, "Solulo", 
             ifelse(dat2008$Community_code ==41607, "Magene", 
             ifelse(dat2008$Community_code ==41608, "Murateia", 
             ifelse(dat2008$Community_code ==41609, "Namaquete", 
             ifelse(dat2008$Community_code ==31101, "Mtopa", 
             ifelse(dat2008$Community_code ==31102, "Hori", 
             ifelse(dat2008$Community_code ==31103, "Mecane (Oua)", 
             ifelse(dat2008$Community_code ==31104, "Nacocolo", 
             ifelse(dat2008$Community_code ==31105, "Pacone", 
             ifelse(dat2008$Community_code ==31106, "Ampuitine",
             ifelse(dat2008$Community_code ==31107, "Corane",
             ifelse(dat2008$Community_code ==31108, "Lalaue",
             ifelse(dat2008$Community_code ==31109, "Mavele",
             ifelse(dat2008$Community_code ==31110, "Mucoroma",
             ifelse(dat2008$Community_code ==31111, "Nacalela",
             ifelse(dat2008$Community_code ==31112, "Nambui",
             ifelse(dat2008$Community_code ==31113, "Namichir",
             ifelse(dat2008$Community_code ==31114, "Nanheua",
             ifelse(dat2008$Community_code ==31115, "Natere",
             ifelse(dat2008$Community_code ==31116, "Tipane",
             ifelse(dat2008$Community_code ==30101, "Iarupa",
             ifelse(dat2008$Community_code ==30102, "Mituco",
             ifelse(dat2008$Community_code ==30103, "Macogone",
             ifelse(dat2008$Community_code ==30104, "Munar",
             ifelse(dat2008$Community_code ==30105, "Namiepe",
             ifelse(dat2008$Community_code ==30106, "Nauluco",
             ifelse(dat2008$Community_code ==30107, "Pulizica",
             ifelse(dat2008$Community_code ==30108, "Naheco",
             ifelse(dat2008$Community_code ==30109, "Namame",NA))))))))))))))))))))))))))))))))))

After:

Community2008 <- case_when(
  dat2008$Community_code == 241601 ~ "Fuzi",
  dat2008$Community_code == 41602 ~ "Moebase",
  dat2008$Community_code==41603 ~ "Txotxo",
  dat2008$Community_code ==41604 ~ "Monudo/Murudo",
  dat2008$Community_code ==41605 ~ "Namige",
  dat2008$Community_code ==41606 ~ "Solulo",
  dat2008$Community_code ==41607 ~ "Magene",
  dat2008$Community_code ==41608 ~ "Murateia",
  dat2008$Community_code ==41609 ~ "Namaquete",
  dat2008$Community_code ==31101 ~ "Mtopa",
  dat2008$Community_code ==31102 ~ "Hori",
  dat2008$Community_code ==31103 ~ "Mecane (Oua)",
  dat2008$Community_code ==31104 ~ "Nacocolo",
  dat2008$Community_code ==31105 ~ "Pacone",
  dat2008$Community_code ==31106 ~ "Ampuitine",
  dat2008$Community_code ==31107 ~ "Corane",
  dat2008$Community_code ==31108 ~ "Lalaue",
  dat2008$Community_code ==31109 ~ "Mavele",
  dat2008$Community_code ==31110 ~ "Mucoroma",
  dat2008$Community_code ==31111 ~ "Nacalela",
  dat2008$Community_code ==31112 ~ "Nambui",
  dat2008$Community_code ==31113 ~ "Namichir",
  dat2008$Community_code ==31114 ~ "Nanheua",
  dat2008$Community_code ==31115 ~ "Natere",
  dat2008$Community_code ==31116 ~ "Tipane",
  dat2008$Community_code ==30101 ~ "Iarupa",
  dat2008$Community_code ==30102 ~ "Mituco",
  dat2008$Community_code ==30103 ~ "Macogone",
  dat2008$Community_code ==30104 ~ "Munar",
  dat2008$Community_code ==30105 ~ "Namiepe",
  dat2008$Community_code ==30106 ~ "Nauluco",
  dat2008$Community_code ==30107 ~ "Pulizica",
  dat2008$Community_code ==30108 ~ "Naheco",
  dat2008$Community_code ==30109 ~ "Namame",
  TRUE ~ NA
)

TODO

Convert multiple ifelse statements that span multiple lines (e.g., https://github.com/Anis-Bensaid/SBT13/blob/master/Script%20R%20Final.R)

License

MIT