[ez][testing] file report: route to pytorch-dev-infra team, remove unused var, etc

.github/workflows/update_test_file_report.yml

@@ -11,7 +11,7 @@ on:
 jobs:
   file-report:
     if: >
-      (github.event_name == 'schedule' && github.event.schedule.cron == '0 2 * * *') ||
+      (github.event_name == 'schedule' && github.event.schedule == '0 2 * * *') ||
       (github.event_name == 'workflow_dispatch')
     permissions:
       issues: write
@@ -25,7 +25,6 @@ jobs:
           cd tools/torchci && python3 -mpip install -e .
 
       - name: Run file report generator
-        if: false
         env:
           CLICKHOUSE_ENDPOINT: ${{ secrets.CLICKHOUSE_HUD_USER_URL }}
           CLICKHOUSE_USERNAME: ${{ secrets.CLICKHOUSE_HUD_USER_USERNAME }}
@@ -46,7 +45,7 @@ jobs:
 
   weekly-file-report-notification:
     if: >
-      (github.event_name == 'schedule' && github.event.schedule.cron == '0 3 * * 2') ||
+      (github.event_name == 'schedule' && github.event.schedule == '0 3 * * 2') ||
       (github.event_name == 'workflow_dispatch')
     permissions:
       issues: write

tools/torchci/test_insights/daily_regression.py

@@ -11,7 +11,7 @@
 
 CONFIG = [
     {
-        "team": "dev-infra",
+        "team": "pytorch-dev-infra",
         "condition": lambda _: True,
         "link": FILE_REPORT_URL,
     },
@@ -74,9 +74,7 @@ def gen_regression_for_team(
             if (info["short_job_name"], info["file"]) in relevant_keys
         ]
 
-        def _sum_invoking_file_info(
-            data: list[dict[str, Any]], field: str
-        ) -> dict[str, Any]:
+        def _sum_invoking_file_info(data: list[dict[str, Any]]) -> dict[str, Any]:
             info = {
                 "count": sum(item["count"] for item in data),
                 "cost": sum(item["cost"] for item in data),
@@ -85,12 +83,8 @@ def _sum_invoking_file_info(
             }
             return info
 
-        agg_prev_file_info = _sum_invoking_file_info(
-            relevant_prev_invoking_file_info, "prev"
-        )
-        agg_curr_file_info = _sum_invoking_file_info(
-            relevant_curr_invoking_file_info, "curr"
-        )
+        agg_prev_file_info = _sum_invoking_file_info(relevant_prev_invoking_file_info)
+        agg_curr_file_info = _sum_invoking_file_info(relevant_curr_invoking_file_info)
 
         invoking_file_info_diff = {
             "count": {

tools/torchci/test_insights/file_report_generator.py

@@ -209,6 +209,7 @@ def _get_frequency(self) -> List[Dict[str, Any]]:
         where
             j.created_at > now() - interval 8 day
             and j.created_at < now() - interval 1 day
+            and j.conclusion != 'cancelled'
         group by
             name
         """

torchci/pages/tests/fileReport.tsx

@@ -464,44 +464,33 @@ function Overview({
   );
 
   const groupedRows = _.map(groupByTarget, (rows, key) => {
-    // Sum within sha
-    const summedBySha = _.map(_.groupBy(rows, "sha"), (shaRows) => {
-      return _.reduce(
-        shaRows,
-        (acc, row) => {
-          acc.count += row.count || 0;
-          acc.time += row.time || 0;
-          acc.cost += row.cost || 0;
-          acc.skipped += row.skipped || 0;
-          acc.frequency += row.frequency || 0;
-          return acc;
-        },
-        { count: 0, time: 0, cost: 0, skipped: 0, frequency: 0 }
-      );
-    });
-    // the reduce across shas for average
-    return _.reduce(
-      summedBySha,
-      (acc, summed) => {
-        acc.count += summed.count;
-        acc.time += summed.time;
-        acc.cost += summed.cost;
-        acc.skipped += summed.skipped;
-        acc.frequency += summed.frequency;
+    // Sum
+    const summed = _.reduce(
+      rows,
+      (acc, row) => {
+        acc.count += row.count || 0;
+        acc.time += row.time || 0;
+        acc.cost += row.cost || 0;
+        acc.skipped += row.skipped || 0;
+        acc.frequency += row.frequency || 0;
         return acc;
       },
-      {
-        id: rows[0].id,
-        file: rows[0].file,
-        short_job_name: rows[0].short_job_name,
-        labels: key,
-        count: 0,
-        time: 0,
-        cost: 0,
-        skipped: 0,
-        frequency: 0,
-      }
+      { count: 0, time: 0, cost: 0, skipped: 0, frequency: 0 }
     );
+
+    // Average across sha data points
+    const numShas = _.uniq(rows.map((r) => r.sha)).length;
+    return {
+      id: rows[0].id,
+      file: rows[0].file,
+      short_job_name: rows[0].short_job_name,
+      labels: key,
+      count: summed.count / numShas,
+      time: summed.time / numShas,
+      cost: summed.cost / numShas,
+      skipped: summed.skipped / numShas,
+      frequency: summed.frequency / numShas,
+    };
   });
 
   return (
@@ -1069,14 +1058,29 @@ export default function Page() {
       <CommitInfo data={data} />
       <Overview
         data={data}
-        setFileFilter={setFileFilter}
-        setJobFilter={setJobFilter}
-        setLabelFilter={setLabelFilter}
+        setFileFilter={(input) => {
+          setFileFilter(input);
+          setFileRegex(false);
+        }}
+        setJobFilter={(input) => {
+          setJobFilter(input);
+          setJobRegex(false);
+        }}
+        setLabelFilter={(input) => {
+          setLabelFilter(input);
+          setLabelRegex(false);
+        }}
       />
       <Diffs
         data={data}
-        setFileFilter={setFileFilter}
-        setJobFilter={setJobFilter}
+        setFileFilter={(input) => {
+          setFileFilter(input);
+          setFileRegex(false);
+        }}
+        setJobFilter={(input) => {
+          setJobFilter(input);
+          setJobRegex(false);
+        }}
       />
       <Graphs data={data} />
       <Stack spacing={2}>