Roy Wedge

Hi @dehiker To test the dashboard, you could try creating a distributed cluster directly and see if that works ``` from distributed import LocalCluster cluster = LocalCluster() cluster ``` reference...

@DManowitz you can set log levels when featuretools is imported using environmental variables `export FEATURETOOLS_LOG_LEVEL=debug` should let you see the stack trace thanks for making the issue, this should be...

> We need to make sure series is set before we get the function in the `feature_set_calculator` What if near the start of `calculate_feature_matrix` we looped through the features and...

This could probably be achieved by updating `AddNumericScalar` and similar primitives to support `Numeric` or `Datetime` inputs

Why is `primitives.base` being renamed to `primitives.core`?

The idea behind `primitives.standard` is that if a primitive was in that module, it was one of the primitives that was part of the basic featuretools installation. Since `featuretools.primitives` loads...

@guilhermemauro , One workaround discussed in #486 is to use `tz_convert` to convert the datetimes to not be timezone aware ```python ct["time"] = ct["time"].dt.tz_convert(None) ```

We can look into adding tz-aware datetimes to our `PandasTypes._pandas_datetimes` list. Would need to think about handling the inability to compare tz-aware and tz-naive datetimes.

When we make this change, we can also move `pyyaml` out of the general requirements. Currently `dask`, `distributed`, and `moto` depend on `pyyaml`. We could make `pyyaml-requirements.txt` and have both...

Perhaps if a categorical column is not flagged as ordered, we sort the category list so we can avoid Pandas not comparing categorical series with the same set of categories?