BUG: fix mutation of DTI backing Series/DataFrame #24096

jbrockmendel · 2018-12-04T17:51:54Z

Fixes (at least some) verify_integrity bugs discussed in #24074.

closes pd.Series(dti) shares data with dti --> DatetimeIndex mutated #21907
tests added / passed
passes git diff upstream/master -u -- "*.py" | flake8 --diff
whatsnew entry

pep8speaks · 2018-12-04T17:51:59Z

Hello @jbrockmendel! Thanks for submitting the PR.

There are no PEP8 issues in the file pandas/core/internals/blocks.py !
There are no PEP8 issues in the file pandas/tests/frame/test_block_internals.py !

jbrockmendel · 2018-12-04T17:53:04Z

pandas/tests/frame/test_block_internals.py

+        assert ser._values is not dti
+        assert dti[1] == ts
+
+    def test_dt64tz_setitem_does_not_mutate_dti(self):


there is not a series.test_block_internals; not sure if there is somewhere else this might belong

pls create one

mroeschke · 2018-12-04T18:46:07Z

Does this fix #6326 as well?

jbrockmendel · 2018-12-04T18:55:59Z

Does this fix #6326 as well?

It looks like it does not. This also breaks another copy/view-based test, which I'll need to look into.

codecov · 2018-12-04T19:54:58Z

Codecov Report

Merging #24096 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #24096      +/-   ##
==========================================
+ Coverage   42.52%   42.52%   +<.01%     
==========================================
  Files         161      161              
  Lines       51697    51699       +2     
==========================================
+ Hits        21982    21986       +4     
+ Misses      29715    29713       -2

Flag	Coverage Δ
#single	`42.52% <100%> (ø)`	⬆️

Impacted Files	Coverage Δ
pandas/core/internals/blocks.py	`52.93% <100%> (+0.19%)`	⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1d3ed91...841c4a5. Read the comment docs.

codecov · 2018-12-04T19:54:58Z

Codecov Report

Merging #24096 into master will decrease coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #24096      +/-   ##
==========================================
- Coverage    92.2%    92.2%   -0.01%     
==========================================
  Files         162      162              
  Lines       51714    51719       +5     
==========================================
+ Hits        47682    47686       +4     
- Misses       4032     4033       +1

Flag	Coverage Δ
#multiple	`90.6% <100%> (ø)`	⬆️
#single	`43.02% <100%> (-0.01%)`	⬇️

Impacted Files	Coverage Δ
pandas/core/internals/blocks.py	`93.72% <100%> (ø)`	⬆️
pandas/core/series.py	`93.7% <100%> (+0.01%)`	⬆️
pandas/core/internals/construction.py	`96.64% <100%> (+0.02%)`	⬆️
pandas/io/json/json.py	`92.61% <0%> (-0.48%)`	⬇️
pandas/util/testing.py	`87.51% <0%> (+0.09%)`	⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9f2c716...4851ffa. Read the comment docs.

…validate

jreback

i don't see any tests for the OP issue here (just the freq change), which AFIACT is orthogonal

jreback · 2018-12-05T12:32:46Z

pandas/core/frame.py

@@ -494,6 +495,11 @@ def _init_dict(self, data, index, columns, dtype=None):
                arrays.loc[missing] = [v] * missing.sum()

        else:
+            for key in data:
+                if isinstance(data[key], ABCIndexClass):


really? we need to deep copy everything?

I can be more specific and just make this ABCDatetimeIndex, but I don't see any solution that doesn't involve making a copy here.

The first thing I tried was to make a copy in Block.setitem (circa line 923 in internals.blocks) if we had a DatetimeIndex. But that breaks a test because it means when we call ser.__setitem__ (or ser.loc.__setitem__, or...), we fail to change other Series that are views on ser.

AFAICT the only way to do do this is to avoid ever having a DatetimeIndex and a Series/column share data. This may be simpler to resolve after Blocks are backed by DatetimeArray instead of DatetimeIndex (I'm not 100% sure on this), but ATM this is a blocker to #24074, which is itself a blocker to DatetimeArray. We could avoid having this be a blocker for #24074 if you were OK with the pickle workaround currently there.

no, the pickle workaround is much worse. for now let's just do this for DTI w/tz only. otherwise this may be a lot of copying. Pls add a TODO around this though.

jreback · 2018-12-05T12:33:25Z

pandas/core/series.py

@@ -189,6 +190,9 @@ def __init__(self, data=None, index=None, dtype=None, name=None,
                else:
                    # need to copy to avoid aliasing issues
                    data = data._values.copy()
+                    if isinstance(data, ABCDatetimeIndex):


really?? this is a specific problem to Datetime w/tz, this is too heavy handed

jreback · 2018-12-05T12:34:37Z

pandas/tests/frame/test_block_internals.py

+        assert ser._values is not dti
+        assert dti[1] == ts
+
+    def test_dt64tz_setitem_does_not_mutate_dti(self):


pls create one

…validate

jbrockmendel · 2018-12-05T18:15:50Z

i don't see any tests for the OP issue here (just the freq change), which AFIACT is orthogonal

Each of the tests has an assertion near the end assert dti[1] == ts or assert dti[0] == ts that is a check for non-mutation.

jreback · 2018-12-05T19:03:32Z

lgtm. ping on green.

jbrockmendel · 2018-12-05T19:52:15Z

Ping

jreback · 2018-12-05T20:19:49Z

thanks!

commit 28c61d770f6dfca6857fd0fa6979d4119a31129e Author: Tom Augspurger <tom.w.augspurger@gmail.com> Date: Thu Dec 6 12:18:19 2018 -0600 uncomment commit bae2e322523efc73a1344464f51611e2dc555ccb Author: Tom Augspurger <tom.w.augspurger@gmail.com> Date: Thu Dec 6 12:17:09 2018 -0600 maybe fixes commit 6cb4db05c9d6ceba3794096f0172cae5ed5f6019 Author: Tom Augspurger <tom.w.augspurger@gmail.com> Date: Thu Dec 6 09:57:37 2018 -0600 we back commit d97ab57fb32cb23371169d9ed659ccfac34cfe45 Merge: a117de4 b78aa8d Author: Tom Augspurger <tom.w.augspurger@gmail.com> Date: Thu Dec 6 09:51:51 2018 -0600 Merge remote-tracking branch 'upstream/master' into disown-tz-only-rebased2 commit b78aa8d Author: gfyoung <gfyoung17+GitHub@gmail.com> Date: Thu Dec 6 07:18:44 2018 -0500 REF/TST: Add pytest idiom to reshape/test_tile (pandas-dev#24107) commit 2993b8e Author: gfyoung <gfyoung17+GitHub@gmail.com> Date: Thu Dec 6 07:17:55 2018 -0500 REF/TST: Add more pytest idiom to scalar/test_nat (pandas-dev#24120) commit b841374 Author: evangelineliu <hsiyinliu@gmail.com> Date: Wed Dec 5 18:21:46 2018 -0500 BUG: Fix concat series loss of timezone (pandas-dev#24027) commit 4ae63aa Author: jbrockmendel <jbrockmendel@gmail.com> Date: Wed Dec 5 14:44:50 2018 -0800 Implement DatetimeArray._from_sequence (pandas-dev#24074) commit 2643721 Author: jbrockmendel <jbrockmendel@gmail.com> Date: Wed Dec 5 14:43:45 2018 -0800 CLN: Follow-up to pandas-dev#24100 (pandas-dev#24116) commit 8ea7744 Author: chris-b1 <cbartak@gmail.com> Date: Wed Dec 5 14:21:23 2018 -0600 PERF: ascii c string functions (pandas-dev#23981) commit cb862e4 Author: jbrockmendel <jbrockmendel@gmail.com> Date: Wed Dec 5 12:19:46 2018 -0800 BUG: fix mutation of DTI backing Series/DataFrame (pandas-dev#24096) commit aead29b Author: topper-123 <contribute@tensortable.com> Date: Wed Dec 5 19:06:00 2018 +0000 API: rename MultiIndex.labels to MultiIndex.codes (pandas-dev#23752)

jbrockmendel added 2 commits December 4, 2018 09:48

fix mutation of DTI backing Series/DataFrame

36a251e

typo fixup

03f4eb4

jbrockmendel commented Dec 4, 2018

View reviewed changes

GH references

616aaf7

jbrockmendel mentioned this pull request Dec 4, 2018

Implement DatetimeArray._from_sequence #24074

Merged

jreback mentioned this pull request Dec 4, 2018

BUG: Series construction sharing data with DTI #22113

Closed

only copy at init-time

841c4a5

jbrockmendel mentioned this pull request Dec 4, 2018

REF: Separate out DataFrame/Series Construction Helpers #24100

Merged

Merge branch 'master' of https://github.com/pandas-dev/pandas into in…

22b32cf

…validate

jreback added the Timeseries label Dec 5, 2018

jreback added this to the 0.24.0 milestone Dec 5, 2018

jreback requested changes Dec 5, 2018

View reviewed changes

jbrockmendel added 2 commits December 5, 2018 08:47

Merge branch 'master' of https://github.com/pandas-dev/pandas into in…

a9e7e40

…validate

implement test_block_internals for series

a66ee4e

only copy if tz is not None, TODO comments

4851ffa

jreback approved these changes Dec 5, 2018

View reviewed changes

jreback merged commit cb862e4 into pandas-dev:master Dec 5, 2018

jbrockmendel deleted the invalidate branch December 5, 2018 20:29

Pingviinituutti pushed a commit to Pingviinituutti/pandas that referenced this pull request Feb 28, 2019

BUG: fix mutation of DTI backing Series/DataFrame (pandas-dev#24096)

df65359

Pingviinituutti pushed a commit to Pingviinituutti/pandas that referenced this pull request Feb 28, 2019

BUG: fix mutation of DTI backing Series/DataFrame (pandas-dev#24096)

1f5c168

jorisvandenbossche mentioned this pull request Jan 14, 2022

BUG: ArrayManager not respecting copy keyword #44889

Merged

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

BUG: fix mutation of DTI backing Series/DataFrame #24096

BUG: fix mutation of DTI backing Series/DataFrame #24096

jbrockmendel commented Dec 4, 2018

pep8speaks commented Dec 4, 2018

jbrockmendel Dec 4, 2018

jreback Dec 5, 2018

jbrockmendel Dec 5, 2018

mroeschke commented Dec 4, 2018

jbrockmendel commented Dec 4, 2018

codecov bot commented Dec 4, 2018

codecov bot commented Dec 4, 2018 •

edited

jreback left a comment

jreback Dec 5, 2018

jbrockmendel Dec 5, 2018

jreback Dec 5, 2018

jreback Dec 5, 2018

jreback Dec 5, 2018

jbrockmendel commented Dec 5, 2018

jreback commented Dec 5, 2018

jbrockmendel commented Dec 5, 2018

jreback commented Dec 5, 2018

BUG: fix mutation of DTI backing Series/DataFrame #24096

BUG: fix mutation of DTI backing Series/DataFrame #24096

Conversation

jbrockmendel commented Dec 4, 2018

pep8speaks commented Dec 4, 2018

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mroeschke commented Dec 4, 2018

jbrockmendel commented Dec 4, 2018

codecov bot commented Dec 4, 2018

Codecov Report

codecov bot commented Dec 4, 2018 • edited

Codecov Report

jreback left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jbrockmendel commented Dec 5, 2018

jreback commented Dec 5, 2018

jbrockmendel commented Dec 5, 2018

jreback commented Dec 5, 2018

codecov bot commented Dec 4, 2018 •

edited