[DataFrame] Added Implementations for equals, query, and some other operations #1610
Conversation
|
Test PASSed. |
| @@ -1224,10 +1248,33 @@ def test_quantile(): | |||
|
|
|||
|
|
|||
| def test_query(): | |||
| ray_df = create_test_dataframe() | |||
| def generate_data(): | |||
There was a problem hiding this comment.
I prefer these be in @pytest.fixture calls and pass the different dataframes to it.
There was a problem hiding this comment.
You can also pass in what you want to query for.
python/ray/dataframe/dataframe.py
Outdated
| @@ -192,6 +201,15 @@ def _map_partitions(self, func, index=None): | |||
|
|
|||
| return DataFrame(new_df, self.columns, index=index) | |||
|
|
|||
| def _update_inplace(self, new_dfs): | |||
There was a problem hiding this comment.
If we're adding this convenience method (which I agree is a good idea) we will also need to update columns and index.
There was a problem hiding this comment.
I added them, and you can update just the fields you want.
python/ray/dataframe/dataframe.py
Outdated
| @@ -72,6 +75,12 @@ def _default_index(self): | |||
|
|
|||
| index = property(_get_index, _set_index) | |||
|
|
|||
| def _update_lengths(self): | |||
There was a problem hiding this comment.
Prefer name change to _compute_lengths(). We have a set and the name doesn't sound much different.
python/ray/dataframe/dataframe.py
Outdated
| def _update_inplace(self, new_dfs): | ||
| """Updates the current DataFrame inplace | ||
| """ | ||
| assert(len(new_dfs) > 0) |
There was a problem hiding this comment.
DataFrames can be empty. This is fine to leave for now, but we need to handle this case (also above in the constructor)
There was a problem hiding this comment.
I agree, I just conformed to the current constructor for now, but this should be changed later.
kunalgosar
force-pushed
the
dataframes
branch
from
3b45aef to
bfbd34e
Compare
|
Test PASSed. |
|
Test PASSed. |
|
Passed on private-travis. OK to merge. Thanks @kunalgosar! |
What do these changes do?
The APIs for these implementations conform to that of pandas, many of the other operators rely on mapped calls to pandas functions.
This PR implements: