https://github.com/leeper/lookfor
R port of Stata's lookfor command
https://github.com/leeper/lookfor
Last synced: about 1 year ago
JSON representation
R port of Stata's lookfor command
- Host: GitHub
- URL: https://github.com/leeper/lookfor
- Owner: leeper
- Created: 2015-03-09T19:00:50.000Z (over 11 years ago)
- Default Branch: master
- Last Pushed: 2022-10-31T17:08:55.000Z (over 3 years ago)
- Last Synced: 2025-03-24T06:11:52.358Z (over 1 year ago)
- Language: R
- Size: 46.9 KB
- Stars: 7
- Watchers: 1
- Forks: 1
- Open Issues: 1
-
Metadata Files:
- Readme: README.Rmd
Awesome Lists containing this project
README
# lookfor: Find what you're looking for #
**lookfor** is an R port of Stata's `lookfor` command, which helps you find stuff you're looking for inside a dataset (e.g., variables, variable values, variable labels, etc.).
## Package Installation ##
[](http://cran.r-project.org/web/packages/lookfor/)
[](https://travis-ci.org/leeper/lookfor)
[](http://cran.r-project.org/web/packages/lookfor/)
[](http://codecov.io/github/leeper/lookfor?branch=master)
[](http://www.repostatus.org/#wip)
The latest development version on GitHub can be installed using **devtools**:
```R
if(!require("devtools")){
install.packages("devtools")
library("devtools")
}
install_github("leeper/lookfor")
```
## Examples ##
Stata's `lookfor` command searches variables and variable labels for an exact matching text string:
```
. webuse auto
(1978 Automobile Data)
. lookfor weight
storage display value
variable name type format label variable label
---------------------------------------------------------------------
weight int %8.0gc Weight (lbs.)
. lookfor ft
storage display value
variable name type format label variable label
---------------------------------------------------------------------
trunk int %8.0g Trunk space (cu. ft.)
turn int %8.0g Turn Circle (ft.)
```
This is helpful and something easily replicated by searching within the `names` attribute of a data.frame. But R allows multiple objects to be present at one time, meaning that it would be time consuming to `grep` the names of all data.frames that might be active in memory at any given point in time. Similarly, those R objects are also themselves named, have numerous other attributes (`rownames`, `colnames`, `levels`, `class`, etc.), and might be stored within lists or environments. While the `lookfor` command in Stata is helpful, it's actually solving a relatively simple problem. Porting `lookfor` to R requires a much more robust solution for searching many different features of potentially numerous data objects stored in various parts of the R environment.
The major weakness of Stata's `lookfor` command is that it relies only on a plain text, exact matching search algorithm. This makes it difficult to search for variables or labels that match a pattern (e.g., a pattern that might easily be described by a regular expression). As a result, the R port uses `grepl` to find matches, meaning that regular expressions are used by default when trying to find something and arguments can be passed via `...` to control the behavior of `grepl` (e.g., setting `fixed=TRUE` will match an exact string, and `perl=TRUE` will use Perl-style regular expressions).
### Look in a data.frame ###
```{r, results="hide"}
options(width = 100)
```
A basic use case is to search for an observation or a variable within a dataset.
```{r}
library("lookfor")
data(USArrests)
# look for observation
lookin(USArrests, "Alaska")
# look for variable
lookin(USArrests, "Assault")
```
### Look in an environment ###
Relatedly, it is possible to search for objects within an environment (and values or attributes within those objects).
```{r, results="hold"}
x <- new.env()
x$mtcars <- mtcars
x$cars <- letters[1:10]
x$cards <- 1:5
lookin(x, "car")
```
### Look in a list ###
And the same can be applied to a potentially deeply nested list.
```{r}
m <- lm(mpg ~ cyl, data = mtcars)
lookin(m, "mpg")
```
### Look everywhere ###
Lastly, the `lookfor` command can be used to look anywhere (within `.GlobalEnv`, the R search path, objects within objects on the search path, loaded namespaces, etc.).
```{r}
lookfor("package")
```
### Look using regular expression ###
As noted above, Stata's `lookfor` command can only match exact/partial character strings. The R port by uses `grepl`, meaning that regular expressions are supported by default, thus allowing for complex pattern matching.
```{r}
data(mtcars)
# Look for car names containing letters and numbers (anywhere)
lookfor("[[:alpha:]]+ [[:digit:]]")
# Look for car names containing letters and numbers (in mtcars)
lookin(mtcars, "[[:alpha:]]+ [[:digit:]]")
```
### Search `comment()` values ###
R's `comment()` function is an underutilized feature that allows users to attach hidden comments to R objects. This is intended (see `? comment`) for helping to annotate R objects with relevant metadata. The value of `comment()` for an object is `NULL` unless it has been set, in which case it is usually a character vector of length 1 (or possibly greater). `lookfor` searches `comment()` values automatically, allowing you to make better use of these annotations.
```{r}
m1 <- lm(mpg ~ cyl, data = mtcars)
comment(m1) <- "model using continuous cylinder"
m2 <- lm(mpg ~ factor(cyl), data = mtcars)
comment(m2) <- "model using factor of cylinder"
lookfor("model using", fixed = TRUE)
```