`fillna()` und `dropna()` in Table and Column

[dorianheidorn]

Is your feature request related to a problem? Please describe

when handling missing values in tables, a function to handle these values would be useful.

Desired solution

Thinking of pandas it could be similar to Table.dropna(list[column_names: str]) and Table.fillna(column_name: str, replace_value). As well as Column.dropna() and Column.fillna(replace_value).

[lars-reimann]

These transformations can already be done using the current API:

• fillna corresponds to our Imputer with a Constant strategy.
• dropna (columns) can be done with drop_columns and list_columns_with_missing_values on a Table.
• dropna (rows) can be done with filter on Table.

I would definitely object fillna since the Imputer does exactly what you want in a more general way and it can also be applied to a test/validation set. dropna might be useful if we find that the task of dropping columns/rows with missing values is very common and the current solution too cumbersome. But then the name of the method should not be dropna but

• drop_columns_with_missing_values
• drop_rows_with_missing_values

These operations should also only be added to Table, not Column.