Remove df property and change to_pandas_dataframe to to_pandas

[tamargrey]

closes #116
closes #147
closes #148

• removes .df property
• changes to_pandas_dataframe to to_pandas
• makes dataframe a protected attribute by changing to .dataframe
• remove replace_none parameter and don't fillna in the DataTables init

[codecov]

Codecov Report

Merging #146 into main will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #146   +/-   ##
=======================================
  Coverage   99.82%   99.82%           
=======================================
  Files          20       20           
  Lines        1740     1742    +2     
=======================================
+ Hits         1737     1739    +2     
  Misses          3        3           

Impacted Files	Coverage Δ
woodwork/data_table.py	100.00% <100.00%> (+0.38%)	⬆️
woodwork/tests/data_table/test_datatable.py	100.00% <100.00%> (ø)
woodwork/tests/testing_utils/data_table_utils.py	100.00% <100.00%> (ø)
woodwork/tests/conftest.py	90.90% <0.00%> (-9.10%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 466a702...e82cbe5. Read the comment docs.

[thehomebrewnerd]

I'm still thinking about whether to_pandas should return a copy of the dataframe or a reference to the dataframe (same with the .dataframe attribute on the table).

Consider this scenario:

• User creates a data table object that contains an integer column - we set this up with Integer logical type and Int64 dtype
• User accesses the dataframe from this data table to use in other calculations and during this the integer column is modified in place - say by a division operation that converts the integers into floats
• Now, if the user runs dt.types on this DataTable, the LogicalType will still be Integer, but the underlying dataframe columns will have a dtype of float64

We are now out of sync - and the logical type information on the DataTable is not accurate. Should we allow this to happen? Same type of thing could happen if a string column got converted into integers/floats.

[gsheni]

minor sugggestions

[gsheni]

Nice

[gsheni]

lgtm