Looker filter expressions

Note: This page is part of the Retrieve and chart data learning series.

Filter expressions are an advanced way to filter Looker queries, and this page describes how to write them. In the Explore section of Looker you can use them by adding a filter and choosing the matches (advanced) option. They are also used in LookML for elements that take a filter parameter.

When using filter expressions in LookML, you should place the expression in quotation marks (see the filters documentation page for proper use). This is especially important for logical values like NULL.

When using filter expressions in the Explore section of Looker, you do not need to place the expression in quotation marks.

Filter expressions are not supported with access grants. See the access_grant documentation page for more information.

String

Matches in string filters depend on the case_sensitive setting in your model file, and on whether your dialect supports case-sensitivity. For example, if case_sensitive is enabled in your model, the expression FOO% will not match the word "food". If case_sensitive is not enabled, or if your dialect does not support case-sensitivity, the expression FOO% will match the word "food".

Example	Description
FOO	is equal to "FOO", exactly
FOO,BAR	is equal to either "FOO" or "BAR", exactly
%FOO%	contains "FOO", matches "buffoon" and "fast food"
FOO%	starts with "FOO", matches "foolish" and "food" but not "buffoon" or "fast food"
%FOO	ends with "FOO", matches "buffoo" and "fast foo" but not "buffoon" or "fast food"
F%OD	starts with an "F" and ends with "OD", matches "fast food"
EMPTY	string is empty (has zero characters) or is null (no value)
NULL	value is null (when it is used as part of a LookML filter expression, place NULL in quotes, as shown on the filters documentation page)
-FOO	is not equal to "FOO" (is any value except "FOO"), matches "pizza", "trash", "fun" but not "foo"
-FOO,-BAR	is not equal to either "FOO" or "BAR", matches any value except "FOO" and "BAR"
-%FOO%	doesn't contain "FOO", does not match "buffoon" or "fast food"
-FOO%	doesn't start with "FOO", does not match "foolish" or "food"
-%FOO	doesn't end with "FOO", does not match "buffoo" or "fast foo"
-EMPTY	string is not empty (has at least one character)
-NULL	value of column is not null (when it is used as part of a LookML filter expression, place -NULL in quotes, as shown on the filters documentation page)
FOO%,BAR	starts with "FOO" or is "BAR" exactly, matches "food" and matches "bar" but not "barfood"
FOO%,-FOOD	starts with "FOO" but is not "FOOD"
_UF	has any single character followed by "UF", matches "buffoon"

Including special characters in string filters

Note these rules for including special characters in string filters:

• To include ", %, or _, prefix with the escape character, ^. For example: ^", ^%, and ^_
• To include a leading -, escape it as ^-. This is only necessary if the - is the leading character; you do not need to escape - if it is inside the string.
• To include ^, escape it as ^^.
• To include a comma in a regular UI string filter, prefix the comma with a backslash character, \. For example: Santa Cruz\, CA.
• To include a comma with the matches (advanced) option in a filter, prefix the comma with the escape character, ^. For example: Santa Cruz^, CA.
• To include a comma in a filter expression in LookML, prefix with the escape character, ^. For example:

  field: filtered_count {
      type: count
      filters: [city: "Santa Cruz^, CA"]
    }

In some cases you may also want to escape or include special characters in string filters when using custom drills to Looker content. See the Things to consider when linking to Looker content section on the Link parameter page for more information.

Date and Time

Looker date filtering allows for English phrases to be used instead of SQL date functions.

Basic structure of date and time filters

For the following examples:

• {n} is an integer.

• {interval} is a time increment such as hours, days, weeks, or months.

  The phrasing you use determines whether the {interval} will include partial time periods or only complete time periods. For example, the expression 3 days includes the current, partial day as well as the prior two days. The expression 3 days ago for 3 days includes the previous three complete days and excludes the current, partial day. See the Relative Dates section for more information.

• {time} can specify a time formatted as either YYYY-MM-DD HH:MM:SS or YYYY/MM/DD HH:MM:SS, or a date formatted as either YYYY-MM-DD or YYYY/MM/DD. When using the form YYYY-MM-DD, be sure to include both digits for the month and day, for example, 2016-01. Truncating a month or day to a single digit is interpreted as an offset, not a date. For example, 2016-1 is interpreted as 2016 minus one year, or 2015.

These are all the possible combinations of date filters:

Combination	Example	Notes
this {interval}	this month	You can use this week, this month, this quarter, or this year. Note that this day is not supported. If you want to get data from the current day, you can use today.
{n} {interval}	3 days
{n} {interval} ago	3 days ago
{n} {interval} ago for {n} {interval}	3 months ago for 2 days
before {n} {interval} ago	before 3 days ago
before {time}	before 2018-01-01 12:00:00	before is not inclusive of the time you specify. The expression before 2018-01-01 will return data from all dates before 2018-01-01, but it will not return data from 2018-01-01.
after {time}	after 2018-10-05	after is inclusive of the time you specify. So, the expression after 2018-10-05 will return data from 2018-10-05 and all dates later than 2018-10-05.
{time} to {time}	2018-05-18 12:00:00 to 2018-05-18 14:00:00	The initial time value is inclusive. The latter time value is not. So the expression 2018-05-18 12:00:00 to 2018-05-18 14:00:00 will return data with the time "2018-05-18 12:00:00" through "2018-05-18 13:59:59".
this {interval} to {interval}	this year to second	The beginning of each interval is used. For example, the expression this year to second returns data from the beginning of the year the query is run through to the beginning of the second the query is run. this week to day returns data from the beginning of the week the query is run through to the beginning of the day the query is run.
{time} for {n} {interval}	2018-01-01 12:00:00 for 3 days
today	today
yesterday	yesterday
tomorrow	tomorrow
{day of week}	Monday	Specifying a day of week with a Dimension Group Date field returns the most recent date that matches the specified day of week. For example, the expression Dimension Group Date matches (advanced) Monday returns the most recent Monday. You can also use {day of week} with the before and after keywords in this context. For example, the expression Dimension Group Date matches (advanced) after Monday returns the most recent Monday and everything after the most recent Monday. The expression Dimension Group Date matches (advanced) before Monday returns every day before the most recent Monday, but it does not return the most recent Monday. Specifying a day of the week with a Dimension Group Day of Week field returns every day that matches the specified day of week. So the expression Dimension Group Day of Week matches (advanced) Monday returns every Monday.
next {week, month, quarter, fiscal quarter, year, fiscal year}	next week	The next keyword is unique in that it requires one of the intervals listed previously and won't work with other intervals.
{n} {interval} from now	3 days from now
{n} {interval} from now for {n} {interval}	3 days from now for 2 weeks

Date filters can also be combined together:

• To get OR logic: Type multiple conditions into the same filter, separated by commas. For example, today, 7 days ago means "today or 7 days ago".
• To get AND logic: Type your conditions, one by one, into multiple date or time filters. For example, you could put after 2014-01-01 into a Created Date filter, then put before 2 days ago into a Created Time filter. This would mean "January 1st, 2014 and after, and before 2 days ago".

Absolute dates

Absolute date filters use the specific date values to generate query results. These are useful when creating queries for specific date ranges.

Example	Description
2018/05/29	sometime on 2018/05/29
2018/05/10 for 3 days	from 2018/05/10 00:00:00 through 2018/05/12 23:59:59
after 2018/05/10	2018/05/10 00:00:00 and after
before 2018/05/10	before 2018/05/10 00:00:00
2018/05	within the entire month of 2018/05
2018/05 for 2 months	within the entire months of 2018/05 and 2018/06
2018/05/10 05:00 for 5 hours	from 2018/05/10 05:00:00 through 2018/05/10 09:59:59
2018/05/10 for 5 months	from 2018/05/10 00:00:00 through 2018/10/09 23:59:59
2018	entire year of 2018 (2018/01/01 00:00:00 through 2018/12/31 23:59:59)
FY2018	entire fiscal year starting in 2018 (if your Looker developers have specified that your fiscal year starts in April then this is 2018/04/01 00:00 through 2019/03/31 23:59)
FY2018-Q1	first quarter of the fiscal year starting in 2018 (if your Looker developers have specified that your fiscal year starts in April then this is 2018/04/01 00:00:00 through 2018/06/30 23:59:59)

Relative dates

Relative date filters allow you to create queries with rolling date values relative to the current date. These are useful when creating queries that update each time you run the query.

For all the examples below, assume today is Friday, 2018/05/18 18:30:02. In Looker, weeks start on Monday unless you change that setting with week_start_day.

Seconds

Example	Description
1 second	the current second (2018/05/18 18:30:02)
60 seconds	60 seconds ago for 60 seconds (2018/05/18 18:29:02 through 2018/05/18 18:30:01)
60 seconds ago for 1 second	60 seconds ago for 1 second (2018/05/18 18:29:02)

Minutes

Example	Description
1 minute	the current minute (2018/05/18 18:30:00 through 18:30:59)
60 minutes	60 minutes ago for 60 minutes (2018/05/18 17:31:00 through 2018/05/18 18:30:59)
60 minutes ago for 1 minute	60 minutes ago for 1 minute (2018/05/18 17:30:00 through 2018/05/18 17:30:59)

Hours

Example	Description
1 hour	the current hour (2018/05/18 18:00 through 2018/05/18 18:59)
24 hours	the same hour of day that was 24 hours ago for 24 hours (2018/05/17 19:00 through 2018/05/18 18:59)
24 hours ago for 1 hour	the same hour of day that was 24 hours ago for 1 hour (2018/05/17 18:00 until 2018/05/17 18:59)

Days

Example	Description
today	the current day (2018/05/18 00:00 through 2018/05/18 23:59)
2 days	all of yesterday and today (2018/05/17 00:00 through 2018/05/18 23:59)
1 day ago	just yesterday (2018/05/17 00:00 until 2018/05/17 23:59)
7 days ago for 7 days	the last complete 7 days (2018/05/11 00:00 until 2018/05/17 23:59)
today for 7 days	the current day, starting at midnight, for 7 days into the future (2018/05/18 00:00 until 2018/05/24 23:59)
last 3 days	2 days ago through the end of the current day (2018/05/16 00:00 until 2018/05/18 23:59)
7 days from now	7 days in the future (2018/05/25 00:00 until 2018/05/25 23:59)

Weeks

Example	Description
1 week	top of the current week going forward (2018/05/14 00:00 through 2018/05/20 23:59)
this week	top of the current week going forward (2018/05/14 00:00 through 2018/05/20 23:59)
before this week	anytime until the top of this week (before 2018/05/14 00:00)
after this week	anytime after the top of this week (2018/05/14 00:00 and later)
next week	the following Monday going forward 1 week (2018/05/21 00:00 through 2018/05/27 23:59)
2 weeks	a week ago Monday going forward (2018/05/07 00:00 through 2018/05/20 23:59)
last week	synonym for "1 week ago"
1 week ago	a week ago Monday going forward 1 week (2018/05/07 00:00 through 2018/05/13 23:59)

Months

Example	Description
1 month	the current month (2018/05/01 00:00 through 2018/05/31 23:59)
this month	synonym for "0 months ago" (2018/05/01 00:00 through 2018/05/31 23:59)
2 months	the past two months (2018/04/01 00:00 through 2018/05/31 23:59)
last month	all of 2018/04
2 months ago	all of 2018/03
before 2 months ago	all time before 2018/03/01
next month	all of 2018/06
2 months from now	all of 2018/07
6 months from now for 3 months	2018/11 through 2019/01

Quarters

Example	Description
1 quarter	the current quarter (2018/04/01 00:00 through 2018/06/30 23:59)
this quarter	synonym for "0 quarters ago" (2018/04/01 00:00 through 2018/06/30 23:59)
2 quarters	the past two quarters (2018/01/01 00:00 through 2018/06/30 23:59)
last quarter	all of Q1 (2018/01/01 00:00 through 2018/03/31 23:59)
2 quarters ago	all of Q4 of last year (2017/010/01 00:00 through 2017/12/31 23:59)
before 2 quarters ago	all time before Q4 of last year
next quarter	all of the following quarter (2018/07/01 00:00 through 2018/09/30 23:59)
2018-07-01 for 1 quarter	all of Q3 (2018/07/01 00:00 through 2018/09/30 23:59)
2018-Q4	all of Q4 (2018/10/01 00:00 through 2018/12/31 23:59)

If your Looker developers have specified using a fiscal year then you can type fiscal in these expressions to use a fiscal quarter instead of a calendar quarter. For example, you can use last fiscal quarter.

Years

Example	Description
1 year	all of the current year (2018/01/01 00:00 through 2018/12/31 23:59)
this year	all of the current year (2018/01/01 00:00 through 2018/12/31 23:59)
next year	all of the following year (2019/01/01 00:00 through 2019/12/31 23:59)
2 years	the past two years (2017/01/01 00:00 through 2018/12/31 23:59)
last year	all of 2017
2 years ago	all of 2016
before 2 years ago	all time before 2016/01/01 (does not include any days between 2016/01/01 and 2016/05/18)

If your Looker developers have specified using a fiscal year then you can type fiscal in these expressions to use a fiscal year instead of a calendar quarter. For example, you can use last fiscal year.

Boolean

Filtering on true or false type values in Looker requires you to know what type of true or false value you're interacting with.

Example	Description
yes or Yes	field evaluates to true Looker developers: for type: yesno dimensions use lowercase, for filters parameters (like those used in a measure or used in an always_filter) use uppercase
no or No	field evaluates to false Looker developers: for type: yesno dimensions use lowercase, for filters parameters (like those used in a measure or used in an always_filter) use uppercase
TRUE	field contains true (for fields that contain Boolean database values)
FALSE	field contains false (for fields that contain Boolean database values)

Number

Filters on numbers support both natural language expressions (for example, 3 to 10) and relational operators (for example, >20). Looker supports the OR operator to express multiple filter ranges (for example, 3 to 10 OR 30 to 100). The AND operator can be used to express numeric ranges with relational operators (for example, >=3 AND <=10) to specify a range. Filters on numbers can also use algebraic interval notation to filter numeric fields.

The syntax for numeric filter expressions using NOT may not be intuitive. If the first filter condition contains a NOT, and no other filter conditions contains a NOT, then all of the filter conditions will be negated. See the below examples for more information.

Example	Description
5	is exactly 5
NOT 5 <>5 !=5	is any value but exactly 5
1, 3, 5, 7	is one of the values 1, 3, 5, or 7, exactly
NOT 66, 99, 4	is not one of the values 66, 99, or 4, exactly
>1, NOT 2, <100	is greater than 1, is not 2, and is less than 100
NOT >1, 2, <100	is less than or equal to 1, is not 2, and is greater than or equal to 100 (Looker recognizes that this is an impossible condition, and will instead write the SQL `IS NULL`)
5, NOT 6, NOT 7	is 5, is not 6 or 7
5.5 to 10 >=5.5 AND <=10	is 5.5 or greater but also 10 or less
NOT 3 to 80.44 <3 OR >80.44	is less than 3 or greater than 80.44
1 to >=1	is 1 or greater
to 10 <=10	is 10 or less
>10 AND <=20 OR 90	is greater than 10 and less than or equal to 20, or is 90 exactly
>=50 AND <=100 OR >=500 AND <=1000	is between 50 and 100, inclusive, or between 500 and 1000, inclusive
NULL	has no data in it (when it is used as part of a LookML filter expression, place NULL in quotes, as shown on the filters documentation page)
NOT NULL	has some data in it (when it is used as part of a LookML filter expression, place NOT NULL in quotes, as shown on the filters documentation page)
(1, 7)	interpreted as 1 < x < 7 where the endpoints are not included. While this notation resembles an ordered pair, in this context it refers to the interval upon which you are working.
[5, 90]	interpreted as 5 <= x <= 90 where the endpoints are included
(12, 20]	interpreted as 12 < x <= 20 where 12 is not included, but 20 is included
[12, 20)	interpreted as 12 <= x < 20 where 12 is included, but 20 is not included
(500, inf)	interpreted as x > 500 where 500 is not included and infinity is always expressed as being "open" (not included). inf may be omitted and (500, inf) may be written as (500,)
(-inf, 10]	interpreted as x <= 10 where 10 is included and infinity is always expressed as being "open" (not included). inf may be omitted and (-inf, 10] may be written as (,10]
[0,9],[20,29]	the numbers between 0 and 9 inclusive or 20 to 29 inclusive
[0,10],20	0 to 10 inclusive or 20
NOT (3,12)	interpreted as x < 3 and x > 12

Location

Location filter expressions are based on latitude and longitude, but can accept some natural language to define boxes and circles within which to limit a search.

Example	Description
36.97, -122.03	location is exactly at latitude 36.97, longitude 122.03
40 miles from 36.97, -122.03	location is within 40 miles of latitude 36.97, longitude -122.03
inside box from 72.33, -173.14 to 14.39, -61.70	location is within a box whose northwest corner is at latitude 72.33, longitude -173.14, and whose southeast corner is at latitude 14.39, longitude -61.70
NOT NULL (works the same as -NULL)	location has both a non-null latitude and a non-null longitude (when it is used as part of a LookML filter expression, place NOT NULL in quotes, as shown on the filters documentation page)
-NULL (works the same as NOT NULL)	location has both a non-null latitude and a non-null longitude (when it is used as part of a LookML filter expression, place -NULL in quotes, as shown on the filters documentation page)
NULL	location has a null latitude, or a null longitude, or both are null (when it is used as part of a LookML filter expression, place NULL in quotes, as shown on the filters documentation page)

Supported units of measurement

To filter in an area around a certain location, you can use these units:

• Meters
• Feet
• Kilometers
• Miles

Singular units of measurement are not supported. For example, filtering for a one-mile radius should be written within 1 miles of 36.97, -122.03.

User Attribute Values

To use the value of a user attribute in a filter expression, reference the user attribute with the _user_attributes Liquid variable:

{{ _user_attributes['name_of_attribute'] }}

For example, suppose you need to apply an sf_ prefix to the value of the salesforce_username user attribute because that is how the values are stored in your database. To add the prefix to the user attribute value, you can add a matches (advanced) filter on the relevant field and use the _user_attributes Liquid variable in the filter expression as follows:

sf_{{_user_attributes['salesforce_username']}}

You can use the same pattern to insert user attributes into LookML dashboard filters and dashboard element filters.