lbl_list_matching_val_lbls
Title
lbl_list_matching_val_lbls - List value labels whose labels match a pattern.
Syntax
lbl_list_matching_val_lbls, pattern(string) [negate verbose varlist(varlist)]
| options | Description |
|---|---|
| pattern(string) | Pattern to find in an answer option. Provide either a substring or a regular expression. |
| negate | Inverts the search, returning value labels that do __not__match the pattern. |
| verbose | Print out labels that match query. Output corresponds to label list lblnames. |
| varlist(varlist) | Restrict the scope of variables to consider |
Description
While Stata offers some tools for searching the content of variable labels (e.g. lookfor), it does not have any methods for similarly searching the contents of value labels.
This command aims fill this gap by:
- Searching labels in value labels for a pattern
- Identifying variable labels that contain labels of interest
- Compiling variables that have these labels of interest attached
This command can be particularly useful for checking that variable do (not) contain patterns of interest. Consider for example:
- Confirming that value labels contain (e.g., no)
- Identifying value labels that deviate from standards
Options
pattern(string) provides the text pattern to find in the contents of value labels. Rather than the traditional Stata glob pattern, this pattern is a sub-string or a regular expression.
negate inverts the search, returning value labels that do __not__match the pattern. In isolation, pattern("my_text") looks for value labels containing "my_text". With negate, pattern("my_search") search looks instead for value labels that do not contain "my_text".
verbose manages the how much output is printed. If the verbose option is not provided, lbl_list_matching_val_lbls reports on whether any matches were found–and, if so, how many value labels match and how many variables the matching value labels describe. If the verbose option is specified, the command will additionally print the contents of the matching value labels as a convenience.
varlist(varlist) restricts the scope of the search to the user-provided variable list. By default, the command searches for matches in all variables in memory. With varlist(), the scope of the search can be narrowed.
Examples
Example 1: contain a pattern
* create some fake data
gen var1 = .
gen var2 = .
gen var3 = .
gen var4 = .
* create some value labels
label define var1_lbl 1 "Yes" 2 "No"
label define var2_lbl 1 "Oui" 2 "Non" 3 "Oui, oui"
label define var4_lbl 1 "Oui" 2 "Non"
* apply those labels to some, but not all, variables
label values var1 var1_lbl
label values var2 var2_lbl
label values var4 var4_lbl
* find value labels with "Oui" and/or "oui" in at least one constituent label
lbl_list_matching_val_lbls, pattern("[Oo]ui")
* find value labels and print out the contents of the label, for convenience
* i.e., to avoid the next step that many users might logically make:
* [label list matching_lbl]
lbl_list_matching_val_lbls, pattern("[Oo]ui") verboseExample 2: do not contain a pattern
* find value labels that do not contain a certain pattern
* for example, no "Oui"/"oui" in yes/no labels from a French-language survey
lbl_list_matching_val_lbls, pattern("[Oo]ui") negateExample 3: contain only a certain set of characters
* create some value labels
label drop _all
* var1_lbl var2_lbl var4_lbl
label define var1_lbl 1 "YES" 2 "NO"
label define var2_lbl 1 "Yes" 2 "No"
label define var3_lbl 1 "yes" 2 "no"
label define var4_lbl 1 "Où" 2 "Là"
* attach them to variables created above
label values var1 var1_lbl
label values var2 var2_lbl
label values var3 var3_lbl
label values var4 var4_lbl
* contains no lower-case characters
lbl_list_matching_val_lbls, pattern("[:lower:]") negate
* contains no French characters
lbl_list_matching_val_lbls, pattern("[àâäÀÂÄéèêëÉÈÊËîïôöÔÖùûüçÇ]") negate
Feedback, bug reports and contributions
Read more about these commands on this repo where this package is developed. Please provide any feedback by opening an issue. PRs with suggestions for improvements are also greatly appreciated.