fixed conversion to sparse for non-numeric index #11856
Conversation
jreback
added
Sparse
|
this looks reasonable. I think you may have to adjust some tests, and maybe some more fixing of the dtype propogation. |
|
I fixed one test to its original state (accounting for conversion of integer values to floats with |
|
can you rebase when you have a chance. |
DSLituiev
force-pushed
the
sparse_nonnumeric_index
branch
from
8283a33 to
177355e
Compare
|
rebased |
|
need some tests for this. |
|
I added a test. |
| @@ -1720,6 +1720,17 @@ def test_nan_columnname(self): | |||
| nan_colname_sparse = nan_colname.to_sparse() | |||
| self.assertTrue(np.isnan(nan_colname_sparse.columns[0])) | |||
|
|
|||
| def test_pivot(self): | |||
| df = pd.DataFrame( list(zip([3,2,4,1,5,3,2], | |||
There was a problem hiding this comment.
add the issue number as a comment; don't use pd.
DSLituiev
force-pushed
the
sparse_nonnumeric_index
branch
from
852838d to
8bc8f28
Compare
|
The current fix still does not completely solve the problem of sparse pivot #11633 (it uses sparse as input, but produces a dense output). One will need to look into |
|
@DSLituiev that's ok, let's get this bug fix in, then can work on that (which is a bit more complicated). you can always start constructing a fix based on THIS branch if you would like. |
|
@DSLituiev can you update? |
|
@DSLituiev can you update, lots of Sparse changes recently. |
|
I'll have a look On Sat, Apr 9, 2016 at 8:05 AM, Jeff Reback [email protected]
|
DSLituiev
force-pushed
the
sparse_nonnumeric_index
branch
from
8bc8f28 to
990c073
Compare
DSLituiev
force-pushed
the
sparse_nonnumeric_index
branch
from
990c073 to
614fb2f
Compare
|
can you rebase/update |
|
can you rebase / update, we've had lots of sparse changes. |
|
superseded by #13021 |
closes #11633 closes #11856 Author: sinhrks <[email protected]> Closes #13201 from sinhrks/sparse_isnull and squashes the following commits: 443b47e [sinhrks] BUG: Sparse creation with object dtype may raise TypeError
this fixes #11633
this fix breaks some sparse tests,
stuck on
test_combine_first:ValueError: total size of new array must be unchanged