Fine-grained errstate handling #13145
Conversation
|
|
||
| .. _whatsnew_0190.new_features: | ||
|
|
||
| New features |
There was a problem hiding this comment.
just leave all this off for now (this whatsnew file)
There was a problem hiding this comment.
As in don't add the new file at all?
|
@rkern ohh that was easy 😄 can we have a test which asserts the actual error state (is the default). Put in |
jreback
added
the
Compat
Current coverage is 85.27% (diff: 100%)@@ master #13145 diff @@
==========================================
Files 139 139
Lines 50380 50439 +59
Methods 0 0
Messages 0 0
Branches 0 0
==========================================
+ Hits 42953 43012 +59
Misses 7427 7427
Partials 0 0
|
|
Please excuse the CI churn. The jobs for the older commits could be cancelled. |
|
no prob. thats an annoyance about Travis. |
|
@rkern I think we can now do this for 0.19.0 can you rebase / update? |
|
@rkern can you rebase |
|
@rkern, you now have another watcher on this... :-) Do you have an ETA when you could rebase? |
|
@jreback Rebased. |
| @@ -95,7 +95,7 @@ def _align_core(terms): | |||
| term_axis_size = len(ti.axes[axis]) | |||
| reindexer_size = len(reindexer) | |||
|
|
|||
| ordm = np.log10(abs(reindexer_size - term_axis_size)) | |||
| ordm = np.log10(max(1, abs(reindexer_size - term_axis_size))) | |||
There was a problem hiding this comment.
this is the fix for #13135 ?
|
this looks pretty good. just some minor comments. can you add in the whatsnew a sub-section, then add a warning (so its a big red box!) near the top. |
|
can you add docs and rebase. ping when green. this touches lots of code so I will merge then. |
…ed to be silenced.
|
@jreback rebased and docced. |
|
looks fine. though failing on all sparse comparisons....pls ping when green |
|
@jreback Fixed thinko. All green. |
|
@rkern Thanks a lot! |
git diff upstream/master | flake8 --diffThe precise strategy to be taken here is open for discussion. I tried to be reasonably fine-grained rather than slap a generic decorator over everything because it's easier to go that direction than the reverse. The
errstate()blocks in the tests were added after fixing all of the library code. Unfortunately, these are less fine-grained than I would like because some of the tests have many lines of the formassert_array_equal(pandas_expression_to_test, expected_raw_numpy_expression)whereexpected_raw_numpy_expressionis what is triggering the warning. It was tedious to try to rewrite all of that to wrap justexpected_raw_numpy_expression.I think I got everything exercised by the test suite except for parts of the test suite that are skipped on my machine due to dependencies. We'll see how things go in the CI.
I haven't added any new tests yet. Could do if requested.