feat(arithmetic): ✖️ multiplication #1397

cmp0xff · 2025-09-30T21:24:17Z

Simplifies exisinting __mul__, mul, etc. in Series
Addresses Index arithmetic overloads show wrong types #1347 and Create tests for arithmetic for TimedeltaIndex and DatetimeIndex #1378
Tests added

Dr-Irv · 2025-10-01T14:32:43Z

Will await your ping for the review.

cmp0xff

@Dr-Irv this #1397 is ready to be reviewed, thanks.

tests/series/test_series.py

cmp0xff · 2025-10-01T21:27:54Z

tests/test_timefuncs.py

-def test_fail_on_adding_two_timestamps() -> None:
-    s1 = pd.Series(pd.to_datetime(["2022-05-01", "2022-06-01"]))
-    s2 = pd.Series(pd.to_datetime(["2022-05-15", "2022-06-15"]))
-    if TYPE_CHECKING_INVALID_USAGE:
-        ssum = s1 + s2  # type: ignore[operator] # pyright: ignore[reportOperatorIssue]
-        ts = pd.Timestamp("2022-06-30")
-        tsum = s1 + ts  # type: ignore[operator] # pyright: ignore[reportOperatorIssue]


Now in tests/series/arithmetic/timestamp/test_add.py

cmp0xff · 2025-10-01T21:29:02Z

tests/test_timefuncs.py

-    std = (df["x"] < df["y"]) * pd.Timedelta(10, "minutes")
-    check(assert_type(std, "pd.Series[pd.Timedelta]"), pd.Series, pd.Timedelta)


tests/series/arithmetic/bool/test_mul.py

cmp0xff · 2025-10-01T21:30:54Z

tests/test_timefuncs.py

-    r3 = td + ts
-    check(assert_type(r3, "pd.Series[pd.Timestamp]"), pd.Series, pd.Timestamp)


tests/series/arithmetic/timedelta/test_add.py

tests/scalars/test_scalars.py

cmp0xff · 2025-10-04T17:15:54Z

@Dr-Irv this #1397 is ready to be reviewed, thanks.

I will mention @Dr-Irv again just in case you missed this pull request. Thanks!

Dr-Irv

Should tests be added that make sure that if a str is used in multiplication, we catch that as invalid? And that both Series[str] and Index[str] with multiplication are caught as well?

This is in pretty good shape. I think you are missing some tests on Index[Any] with multiplication.

Dr-Irv · 2025-10-04T17:13:58Z

tests/indexes/arithmetic/test_mul.py

Not sure why you have an empty file here, although I would assume some tests are necessary???

Added in c8e0105

Dr-Irv · 2025-10-04T17:18:52Z

tests/indexes/arithmetic/timedelta/test_mul.py

+def left() -> "pd.Index[pd.Timedelta]":
+    """left operand"""
+    # pandas-dev/pandas#62524
+    lo = pd.Index([1]) * [timedelta(seconds=1)]  # left operand
+    return check(assert_type(lo, "pd.Index[pd.Timedelta]"), pd.Index, timedelta)


I don't think we should be testing this. pd.Index[Timedelta] should only hold Timedelta objects, not python timedelta objects. pd.Index([timedelta(seconds=1)]) creates this: TimedeltaIndex(['0 days 00:00:01'], dtype='timedelta64[ns]', freq=None) and that's the normal supported usage of timedelta in an index.

I'd even question whether we should have any operations supporting pd.Index[pd.Timedelta], because that really is a TimedeltaIndex .

Actually, thinking about it more, I don't think we should have Index[Timedelta] or Index[Timestamp] in the stubs.

This is in pandas-dev/pandas#62524, where I found out that pandas indeed creates Index[Timedelta] when one does pd.Index([1]) * [timedelta(days=1)], which is quite unintuitive.

Dr-Irv · 2025-10-04T17:47:28Z

@Dr-Irv this #1397 is ready to be reviewed, thanks.

I will mention @Dr-Irv again just in case you missed this pull request. Thanks!

I had to wait until today when I had time to review this large of a PR

cmp0xff

Tests for Series[str] * num and Index[str] * num added in 6cf48fc

cmp0xff · 2025-10-04T19:23:14Z

tests/indexes/arithmetic/test_mul.py

Added in c8e0105

cmp0xff · 2025-10-04T19:24:38Z

tests/indexes/arithmetic/timedelta/test_mul.py

+def left() -> "pd.Index[pd.Timedelta]":
+    """left operand"""
+    # pandas-dev/pandas#62524
+    lo = pd.Index([1]) * [timedelta(seconds=1)]  # left operand
+    return check(assert_type(lo, "pd.Index[pd.Timedelta]"), pd.Index, timedelta)


This is in pandas-dev/pandas#62524, where I found out that pandas indeed creates Index[Timedelta] when one does pd.Index([1]) * [timedelta(days=1)], which is quite unintuitive.

cmp0xff · 2025-10-05T09:58:51Z

/pyright_strict

cmp0xff · 2025-10-05T10:02:12Z

/pandas_nightly

cmp0xff · 2025-10-05T10:02:26Z

/mypy_nightly

cmp0xff · 2025-10-05T10:14:44Z

In pandas_nightly (see the pipeline run):

timedelta64 and str multiplying bool are forbidden. Shall we forbid them now?
check(assert_type(s.info(show_counts=False), None), type(None)) is broken, I think we should report it.

cmp0xff force-pushed the feature/cmp0xff/arithmetic-mul branch from f1acc37 to 19191a4 Compare October 1, 2025 20:22

cmp0xff added 5 commits October 1, 2025 23:21

feat(arithmetic): multiplication

0444358

feat: mul for timedelta indexes

8d2a17f

doc and refactor

b9c57d2

complex

d2d11d1

bool

9c802b4

cmp0xff force-pushed the feature/cmp0xff/arithmetic-mul branch from 19191a4 to 9c802b4 Compare October 1, 2025 21:22

cmp0xff commented Oct 1, 2025

View reviewed changes

cmp0xff marked this pull request as ready for review October 1, 2025 21:37

cmp0xff added 3 commits October 2, 2025 09:53

chore(poetry): update pyright

8b987d4

Merge branch 'main' into feature/cmp0xff/arithmetic-mul

78b009a

Merge branch 'main' into feature/cmp0xff/arithmetic-mul

cda5639

cmp0xff requested a review from Dr-Irv October 4, 2025 17:16

Dr-Irv requested changes Oct 4, 2025

View reviewed changes

cmp0xff added 4 commits October 4, 2025 20:31

fix(comment): pandas-dev#1397 (comment)

c8e0105

fix(comment): pandas-dev#1397 (review)

6cf48fc

chore(pyrefly): remove-unused-ignores

560695e

Merge branch 'main' into feature/cmp0xff/arithmetic-mul

f2b471c

cmp0xff commented Oct 4, 2025

View reviewed changes

cmp0xff requested a review from Dr-Irv October 4, 2025 19:26

		std = (df["x"] < df["y"]) * pd.Timedelta(10, "minutes")
		check(assert_type(std, "pd.Series[pd.Timedelta]"), pd.Series, pd.Timedelta)

		r3 = td + ts
		check(assert_type(r3, "pd.Series[pd.Timestamp]"), pd.Series, pd.Timestamp)

Uh oh!

feat(arithmetic): ✖️ multiplication #1397

Are you sure you want to change the base?

feat(arithmetic): ✖️ multiplication #1397

Conversation

cmp0xff commented Sep 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Dr-Irv commented Oct 1, 2025

Uh oh!

cmp0xff left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cmp0xff commented Oct 4, 2025

Uh oh!

Dr-Irv left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Dr-Irv commented Oct 4, 2025

Uh oh!

cmp0xff left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

cmp0xff commented Oct 5, 2025

Uh oh!

cmp0xff commented Oct 5, 2025

Uh oh!

cmp0xff commented Oct 5, 2025

Uh oh!

cmp0xff commented Oct 5, 2025

Uh oh!

Uh oh!

cmp0xff commented Sep 30, 2025 •

edited

Loading