Add tests for ingestion of files with missing columns

[diskontinuum]

See also pycytominer/issues/79.

[diskontinuum]

Please ignore several involuntary enter-typos for the title above ;)

[diskontinuum]

The current cytominer-database version (only with the --parquet option) deals with missing columns explicitly by aligning dataframes (that will be concatenated and written to a Parquet file) with a specific reference dataframe:
`aligned_df, ref_df = df.align(ref_df, axis=1)``

Missing columns will be added as NaN-valued columns by default (with same column name and type as the reference frame). To choose a specific value, we could add the parameter fill_value to pandas.dataframe.align() .