Han Wang
**Describe the solution you'd like** Currently, there are a lot of duplicated but unnecessary unit tests, taking a lot of time for each build, we should make the test suites...
**Describe the solution you'd like** Currently when saving avro, we lose the original types. This can be done if we implement something like [this](https://github.com/ynqa/pandavro/blob/master/pandavro/__init__.py).
```python # input_has: a # schema: * def tr(df:pd.DataFrame) -> pd.DataFrame: return df dag = FugueWorkflow() dag.df([[0]],"b:int").transform(tr).persist() # this step will not fail because input_has is runtime validation dag.run() #...
Currently, most workflow related exceptions are simply FugueWorkflowError. In the code it can be more explicit. In the future, we can see what runtime error can be caught at compile...
It is possible to analyze the dag and find when to persist and unpersist a df
**Is your feature request related to a problem? Please describe.** Currently there are manual steps in the code to start using dask-sql with Fugue. **Describe the solution you'd like** I...
`GROUP BY` can be implicit for common cases ```SQL SELECT a, b, COUNT(*) AS c GROUP BY a,b ``` Omitting `GROUP BY` doesn't cause ambiguity in this case. So ```SQL...
Fugue needs to support Map type. Map type is in the form of and the data is in the form of a list of key-value tuples or just a dict....