Backport PR #2408 on branch 0.12.x (fix: benchmark timeout for too much outer joins on concat) (#2409)

meeseeksmachine · ilan-gold · web-flow · commit 2c2b2f822ea6 · 2026-04-24T13:26:28.000+02:00
Co-authored-by: Ilan Gold &lt;ilanbassgold@gmail.com&gt;
diff --git a/benchmarks/benchmarks/sparse_dataset.py b/benchmarks/benchmarks/sparse_dataset.py
@@ -107,13 +107,13 @@ def setup(self, *_):
             AnnData(
                 var=pd.DataFrame(
                     index=[
-                        f"gene_{j}{f'_{i}' if (j % 100 == 0) else ''}"
+                        f"gene_{j}{f'_{i}' if (j % 500 == 0) else ''}"
                         for j in range(10_000)
                     ]
                 ),
                 X=read_elem_lazy(self.group["X"]),
             )
-            for i in range(10)
+            for i in range(5)
         ]
 
     def time_concat(self, join: Literal["inner", "outer"], fill_value: Literal[0, -1]):

Original file line number	Diff line number	Diff line change
`@@ -107,13 +107,13 @@ def setup(self, *_):`
`107`	`107`	`AnnData(`
`108`	`108`	`var=pd.DataFrame(`
`109`	`109`	`index=[`
`110`		`- f"gene_{j}{f'_{i}' if (j % 100 == 0) else ''}"`
	`110`	`+ f"gene_{j}{f'_{i}' if (j % 500 == 0) else ''}"`
`111`	`111`	`for j in range(10_000)`
`112`	`112`	`]`
`113`	`113`	`),`
`114`	`114`	`X=read_elem_lazy(self.group["X"]),`
`115`	`115`	`)`
`116`		`- for i in range(10)`
	`116`	`+ for i in range(5)`
`117`	`117`	`]`
`118`	`118`
`119`	`119`	`def time_concat(self, join: Literal["inner", "outer"], fill_value: Literal[0, -1]):`