Enable index operations. #1955
Enable index operations. #1955
Conversation
Codecov Report
@@ Coverage Diff @@
## master #1955 +/- ##
==========================================
- Coverage 94.63% 94.47% -0.17%
==========================================
Files 49 49
Lines 10829 10855 +26
==========================================
+ Hits 10248 10255 +7
- Misses 581 600 +19
Continue to review full report at Codecov.
|
| # only work between at most two `Index`s. We might need to fix it in the future. | ||
| self_len = len(self) | ||
| if any(len(col) != self_len for col in args if isinstance(col, IndexOpsMixin)): | ||
| raise ValueError("operands could not be broadcast together with shapes") |
There was a problem hiding this comment.
How does length comparison relate to broadcast in the error message?
There was a problem hiding this comment.
The term "broadcast" here is not the same as Spark's "broadcast".
Maybe the term should've been "broad-cast", which means cast to broader types, or in this case cast to larger size. I'd just follow pandas' error message.
There was a problem hiding this comment.
I see, maybe that's what we are talking about
>>> z= np.arange(12).reshape(3,4)
>>> m= np.arange(9).reshape(3,3)
>>> z * m
Traceback (most recent call last):
File "<input>", line 1, in <module>
z * m
ValueError: operands could not be broadcast together with shapes (3,4) (3,3)
The current error message also looks good.
| "Cannot combine the series or dataframe because it comes from a different dataframe. " | ||
| "In order to allow this operation, enable 'compute.ops_on_diff_frames' option." | ||
| ) | ||
| raise ValueError(ERROR_MESSAGE_CANNOT_COMBINE) | ||
|
|
||
|
|
||
| def align_diff_frames( |
| @@ -376,12 +385,11 @@ def align_diff_frames( | |||
| return kdf | |||
|
|
|||
|
|
|||
| def align_diff_series(func, this_series, *args, how="full"): | |||
| from databricks.koalas.base import IndexOpsMixin | |||
| def align_diff_series(func, this_series: "Series", *args, how: str = "full") -> "Series": | |||
There was a problem hiding this comment.
I'm just wondering whether we should keep align_diff_series here or move to series.py since it is only for Series.
There was a problem hiding this comment.
Yeah, maybe we can move it, but I'd leave it to the future PRs.
|
Thanks! I'd merge this now. Please feel free to leave comments if any. @HyukjinKwon @itholic |
Currently Koalas can't handle index operations very well.
or
This PR enables those operations: