feature-engineering

Describe the issue:
During computing Channel Dependencies reshape_break_channel_dependency does following code to ensure that the number of input channels equals the number of output channels:

in_shape = op_node.auxiliary['in_shape']
out_shape = op_node.auxiliary['out_shape']
in_channel = in_shape[1]
out_channel = out_shape[1]
return in_channel != out_channel

This is correct

• As a user, I wish featuretools dfs would take a string as cutoff_time aswell as a datetime object

Code Example

fm, features = ft.dfs(entityset=es,
                      target_dataframe_name='customers',
                      cutoff_time="2014-1-1 05:00",
                      instance_ids=[1],
                      cutoff_time_in_index=True)

as well as

Expected Behavior

Feast should allow users to create feature views with .csv data sources and retrieve features from offline store without any issues.

Current Behavior

Presently, I have a .csv file sitting in S3 bucket and I am able to create a feature view using this .csv file but while fetching the features from offline store getting below error

-------------------------

Problem
Some of our transformers & estimators are not thoroughly tested or not tested at all.

Solution
Use OpTransformerSpec and OpEstimatorSpec base test specs to provide tests for all existing transformers & estimators.

When using r2 as eval metric for regression task (with 'Explain' mode) the metric values reported in Leaderboard (at README.md file) are multiplied by -1.
For instance, the metric value for some model shown in the Leaderboard is -0.41, while when clicking the model name leads to the detailed results page - and there the value of r2 is 0.41.
I've noticed that when one of R2 metric values in the L

https://github.com/4paradigm/OpenMLDB/blob/230f99b9b5744f2663b0253a87a2496c9f5ce96b/java/openmldb-taskmanager/src/main/scala/com/_4paradigm/openmldb/taskmanager/JobInfoManager.scala#L141-L170

pstmt is useless, delete it. And delete statement in finally block.

Add missing_only functionality to all imputers to use in combination with variables=None

When variables is None, the imputers select all numerical, or categorical or all variables by default. With the missing_only, it would select only those from each subgroup that show missing data during fit.

In #3324 , we had to mark some tests as expected to fail since XGBoost was throwing a FutureWarning. The warning has been addressed in XGBoost, so we're just waiting for the PR merged to be released. This issue is discussed in the #3275 issue.

evalml/tests/component_tests/test_xgboost_classifier.py needs to have the @pytest.mark.xfail removed f

Is your feature request related to a problem? Please describe.

We should implement useful implementations of:

class ResultMixin(object):
    """Base class housing the static function.

    Why a static function? That's because certain frameworks can only pickle a static function, not an entire
    object.
    """
    @staticmethod
    @abc.abstractmethod
    def build_