Select rows filter

The Select rows filter on the Recipe page allows you to filter rows out of a dataset based on row queries. For example, you can include only the rows where the #org is "UNICEF" or where the #sector is "WASH".

Multiple queries count as a logical OR — a row will be included (or excluded, if inverted) if it matches any of the row queries. If you want a logical AND, include a second Select rows filter with more row queries.

This filter is similar to the SQL "select" command.

Options

Queries: one or more row queries to select rows from the dataset. Use the "+" button to add more queries.

Invert query: if checked, the result will include all the rows that don't match at least one query.

Example

Original data

#org	#sector	#adm1+name	#adm1+code	#targeted
UNICEF	Education	Coast	X001	5000
Save the Children	Education	Plains	X002	300
IOM	CCCM	Coast	X001	1500
UNICEF	Protection	Plains	X002	8000

Sample output

Row query "#sector=education":

#org	#sector	#adm1+name	#adm1+code	#targeted
UNICEF	Education	Coast	X001	5000
Save the Children	Education	Plains	X002	300

Row query "sector!=education":

#org	#sector	#adm1+name	#adm1+code	#targeted
IOM	CCCM	Coast	X001	1500
UNICEF	Protection	Plains	X002	8000

Row query "targeted>2000":

#org	#sector	#adm1+name	#adm1+code	#targeted
UNICEF	Education	Coast	X001	5000
UNICEF	Protection	Plains	X002	8000

Use cases

From a large 3W, extract all the activities for a specific cluster, organisation, country, etc. (e.g. "#sector=WASH")

Choose exceptional (large or small) values (e.g. "#affected>1000000")

Select only activities with a specific status (e.g. "#status=active")

Choose all activities before a specific a #date (e.g. "#date<2016-01-01")

Select only rows for the latest year in the dataset (e.g. "#date+year is max")