feat: Add P75 quantile for Core Web Vitals - and globally

frontend/src/lib/components/InsightLabel/index.tsx

@@ -66,7 +66,7 @@ function MathTag({ math, mathProperty, mathHogQL, mathGroupTypeIndex }: MathTagP
     if (math === 'unique_group' && mathGroupTypeIndex != undefined) {
         return <LemonTag>Unique {aggregationLabel(mathGroupTypeIndex).plural}</LemonTag>
     }
-    if (math && ['sum', 'avg', 'min', 'max', 'median', 'p90', 'p95', 'p99'].includes(math || '')) {
+    if (math && ['sum', 'avg', 'min', 'max', 'median', 'p75', 'p90', 'p95', 'p99'].includes(math)) {
         return (
             <>
                 <LemonTag>{mathDefinitions[math]?.name || capitalizeFirstLetter(math)}</LemonTag>

frontend/src/queries/schema.json

@@ -3906,6 +3906,7 @@
                 "min_count_per_actor",
                 "max_count_per_actor",
                 "median_count_per_actor",
+                "p75_count_per_actor",
                 "p90_count_per_actor",
                 "p95_count_per_actor",
                 "p99_count_per_actor"
@@ -9099,7 +9100,7 @@
             "type": "object"
         },
         "PropertyMathType": {
-            "enum": ["avg", "sum", "min", "max", "median", "p90", "p95", "p99"],
+            "enum": ["avg", "sum", "min", "max", "median", "p75", "p90", "p95", "p99"],
             "type": "string"
         },
         "PropertyOperator": {

frontend/src/scenes/cohorts/CohortFilters/constants.tsx

@@ -89,6 +89,9 @@ export const FIELD_VALUES: Record<FieldOptionsType, FieldValues> = {
             [PropertyMathType.Median]: {
                 label: 'Median',
             },
+            [PropertyMathType.P75]: {
+                label: '75th percentile',
+            },
             [PropertyMathType.P90]: {
                 label: '90th percentile',
             },

frontend/src/scenes/trends/mathsLogic.tsx

@@ -241,6 +241,19 @@ export const PROPERTY_MATH_DEFINITIONS: Record<PropertyMathType, MathDefinition>
         ),
         category: MathCategory.PropertyValue,
     },
+    [PropertyMathType.P75]: {
+        name: '75th percentile',
+        shortName: '75th percentile',
+        description: (
+            <>
+                Event property 75th percentile.
+                <br />
+                <br />
+                For example 100 events captured with property <code>amount</code> equal to 101..200, result in 175.
+            </>
+        ),
+        category: MathCategory.PropertyValue,
+    },
     [PropertyMathType.P90]: {
         name: '90th percentile',
         shortName: '90th percentile',
@@ -315,6 +328,12 @@ export const COUNT_PER_ACTOR_MATH_DEFINITIONS: Record<CountPerActorMathType, Mat
         description: <>Event count per actor 50th percentile.</>,
         category: MathCategory.EventCountPerActor,
     },
+    [CountPerActorMathType.P75]: {
+        name: '75th percentile',
+        shortName: '75th percentile',
+        description: <>Event count per actor 75th percentile.</>,
+        category: MathCategory.EventCountPerActor,
+    },
     [CountPerActorMathType.P90]: {
         name: '90th percentile',
         shortName: '90th percentile',

frontend/src/types.ts

@@ -3651,6 +3651,7 @@ export enum PropertyMathType {
     Minimum = 'min',
     Maximum = 'max',
     Median = 'median',
+    P75 = 'p75',
     P90 = 'p90',
     P95 = 'p95',
     P99 = 'p99',
@@ -3661,6 +3662,7 @@ export enum CountPerActorMathType {
     Minimum = 'min_count_per_actor',
     Maximum = 'max_count_per_actor',
     Median = 'median_count_per_actor',
+    P75 = 'p75_count_per_actor',
     P90 = 'p90_count_per_actor',
     P95 = 'p95_count_per_actor',
     P99 = 'p99_count_per_actor',

posthog/api/documentation.py

@@ -161,6 +161,7 @@ class PersonPropertiesSerializer(serializers.Serializer):
 - `min`: min of a numeric property.
 - `max`: max of a numeric property.
 - `median`: median of a numeric property.
+- `p75`: 75th percentile of a numeric property.
 - `p90`: 90th percentile of a numeric property.
 - `p95` 95th percentile of a numeric property.
 - `p99`: 99th percentile of a numeric property.

posthog/hogql_queries/insights/trends/aggregation_operations.py

@@ -73,6 +73,8 @@ def select_aggregation(self) -> ast.Expr:
                 return self._math_func("max", None)
             elif self.series.math == "median":
                 return self._math_quantile(0.5, None)
+            elif self.series.math == "p75":
+                return self._math_quantile(0.75, None)
             elif self.series.math == "p90":
                 return self._math_quantile(0.9, None)
             elif self.series.math == "p95":
@@ -106,6 +108,7 @@ def is_count_per_actor_variant(self):
             "min_count_per_actor",
             "max_count_per_actor",
             "median_count_per_actor",
+            "p75_count_per_actor",
             "p90_count_per_actor",
             "p95_count_per_actor",
             "p99_count_per_actor",
@@ -247,6 +250,8 @@ def _actors_inner_select_query(
                 math_func = self._math_func("max", ["total"])
             elif self.series.math == "median_count_per_actor":
                 math_func = self._math_quantile(0.5, ["total"])
+            elif self.series.math == "p75_count_per_actor":
+                math_func = self._math_quantile(0.75, ["total"])
             elif self.series.math == "p90_count_per_actor":
                 math_func = self._math_quantile(0.9, ["total"])
             elif self.series.math == "p95_count_per_actor":

posthog/hogql_queries/insights/trends/test/test_aggregation_operations.py

@@ -32,13 +32,15 @@
         [PropertyMathType.MIN, "$browser"],
         [PropertyMathType.MAX, "$browser"],
         [PropertyMathType.MEDIAN, "$browser"],
+        [PropertyMathType.P75, "$browser"],
         [PropertyMathType.P90, "$browser"],
         [PropertyMathType.P95, "$browser"],
         [PropertyMathType.P99, "$browser"],
         [CountPerActorMathType.AVG_COUNT_PER_ACTOR, None],
         [CountPerActorMathType.MIN_COUNT_PER_ACTOR, None],
         [CountPerActorMathType.MAX_COUNT_PER_ACTOR, None],
         [CountPerActorMathType.MEDIAN_COUNT_PER_ACTOR, None],
+        [CountPerActorMathType.P75_COUNT_PER_ACTOR, None],
         [CountPerActorMathType.P90_COUNT_PER_ACTOR, None],
         [CountPerActorMathType.P95_COUNT_PER_ACTOR, None],
         [CountPerActorMathType.P99_COUNT_PER_ACTOR, None],
@@ -79,13 +81,15 @@ def test_all_cases_return(
         [PropertyMathType.MIN, False],
         [PropertyMathType.MAX, False],
         [PropertyMathType.MEDIAN, False],
+        [PropertyMathType.P75, False],
         [PropertyMathType.P90, False],
         [PropertyMathType.P95, False],
         [PropertyMathType.P99, False],
         [CountPerActorMathType.AVG_COUNT_PER_ACTOR, True],
         [CountPerActorMathType.MIN_COUNT_PER_ACTOR, True],
         [CountPerActorMathType.MAX_COUNT_PER_ACTOR, True],
         [CountPerActorMathType.MEDIAN_COUNT_PER_ACTOR, True],
+        [CountPerActorMathType.P75_COUNT_PER_ACTOR, True],
         [CountPerActorMathType.P90_COUNT_PER_ACTOR, True],
         [CountPerActorMathType.P95_COUNT_PER_ACTOR, True],
         [CountPerActorMathType.P99_COUNT_PER_ACTOR, True],

posthog/hogql_queries/insights/trends/test/test_trends.py

@@ -4094,6 +4094,10 @@ def test_max_filtering(self):
     def test_median_filtering(self):
         self._test_math_property_aggregation("median", values=range(101, 201), expected_value=150)
 
+    @also_test_with_materialized_columns(["some_number"])
+    def test_p75_filtering(self):
+        self._test_math_property_aggregation("p75", values=range(101, 201), expected_value=175)
+
     @also_test_with_materialized_columns(["some_number"])
     def test_p90_filtering(self):
         self._test_math_property_aggregation("p90", values=range(101, 201), expected_value=190)

posthog/management/commands/compare_hogql_insights.py

@@ -32,10 +32,10 @@ def handle(self, *args, **options):
         # len(insights)
         for insight in insights[0:500]:
             for event in insight.filters.get("events", []):
-                if event.get("math") in ("median", "p90", "p95", "p99"):
+                if event.get("math") in ("median", "p75", "p90", "p95", "p99"):
                     event["math"] = "sum"
             for event in insight.filters.get("actions", []):
-                if event.get("math") in ("median", "p90", "p95", "p99"):
+                if event.get("math") in ("median", "p75", "p90", "p95", "p99"):
                     event["math"] = "sum"
             try:
                 print(  # noqa: T201

posthog/models/entity/entity.py

@@ -20,24 +20,26 @@
     "monthly_active",
     "unique_group",
     "unique_session",
-    # TODO: When we are finally on Python 3.11+, inline the below as *PROPERTY_MATH_FUNCTIONS.keys()
+    "hogql",
+    # Equivalent to *PROPERTY_MATH_FUNCTIONS.keys(),
     "sum",
     "min",
     "max",
     "avg",
     "median",
+    "p75",
     "p90",
     "p95",
     "p99",
-    # TODO: When we are finally on Python 3.11+, inline the below as *COUNT_PER_ACTOR_MATH_FUNCTIONS.keys()
+    # Equivalent to *COUNT_PER_ACTOR_MATH_FUNCTIONS.keys()
     "min_count_per_actor",
     "max_count_per_actor",
     "avg_count_per_actor",
     "median_count_per_actor",
+    "p75_count_per_actor",
     "p90_count_per_actor",
     "p95_count_per_actor",
     "p99_count_per_actor",
-    "hogql",
 ]
 
 

posthog/queries/test/test_trends.py

@@ -3611,6 +3611,10 @@ def test_max_filtering(self):
     def test_median_filtering(self):
         self._test_math_property_aggregation("median", values=range(101, 201), expected_value=150)
 
+    @also_test_with_materialized_columns(["some_number"])
+    def test_p75_filtering(self):
+        self._test_math_property_aggregation("p75", values=range(101, 201), expected_value=175)
+
     @also_test_with_materialized_columns(["some_number"])
     def test_p90_filtering(self):
         self._test_math_property_aggregation("p90", values=range(101, 201), expected_value=190)

posthog/queries/trends/util.py

@@ -33,6 +33,7 @@
     "min": "min",
     "max": "max",
     "median": "quantile(0.50)",
+    "p75": "quantile(0.75)",
     "p90": "quantile(0.90)",
     "p95": "quantile(0.95)",
     "p99": "quantile(0.99)",
@@ -43,6 +44,7 @@
     "min_count_per_actor": "min",
     "max_count_per_actor": "max",
     "median_count_per_actor": "quantile(0.50)",
+    "p75_count_per_actor": "quantile(0.75)",
     "p90_count_per_actor": "quantile(0.90)",
     "p95_count_per_actor": "quantile(0.95)",
     "p99_count_per_actor": "quantile(0.99)",

posthog/schema.py

@@ -543,6 +543,7 @@ class CountPerActorMathType(StrEnum):
     MIN_COUNT_PER_ACTOR = "min_count_per_actor"
     MAX_COUNT_PER_ACTOR = "max_count_per_actor"
     MEDIAN_COUNT_PER_ACTOR = "median_count_per_actor"
+    P75_COUNT_PER_ACTOR = "p75_count_per_actor"
     P90_COUNT_PER_ACTOR = "p90_count_per_actor"
     P95_COUNT_PER_ACTOR = "p95_count_per_actor"
     P99_COUNT_PER_ACTOR = "p99_count_per_actor"
@@ -1290,6 +1291,7 @@ class PropertyMathType(StrEnum):
     MIN = "min"
     MAX = "max"
     MEDIAN = "median"
+    P75 = "p75"
     P90 = "p90"
     P95 = "p95"
     P99 = "p99"