Universe data frames normalization (#8385)

* Normalize universe data frames

Universe and (generically BaseDataCollection) data frames are not normalize and unpacked into a data frame, instead of just creating data frames with the universe lists within it

* Fix unit tests and algorithms to expecte new universe dataframe format

* Fixes

* Add PandasConverter.DataFrameGenerator class

* Pandas data frame generator class fixes

* Add comments

* Housekeeping

* Add attributes to mark classes and properties for pandas processing

* Improve pandas properties expanding

Allow and handle duplicate names

* Use PandasData generalization for Lean common data types

* Add points time as column when converting base data collections to data frames

* Cleanup and minor changes

* Minor change

* Pandas data to get type members on demand

* Move Pandas helper classes to their own files

* Minor changes

* Add flatten argument to python history api

This allows users to decide whether they want fully expanded dataframes for universe and other collection data types. Else, master behavior is kept

* Adding missing changes to last commit

* Update Pythonnet version to 2.0.40

* Add flattent argument to algorithm's OptionChain api

* Minor changes

* Housekeeping

* Minor changes

* Bug fix skipping data collection data points

* Add comment

* Set correct exchange time to OptionUniverse instances

* Address peer review and cleanup

* Cleanup

* Minor changes

Algorithm.Python/OptionChainFullDataRegressionAlgorithm.py

@@ -28,7 +28,7 @@ class OptionChainFullDataRegressionAlgorithm(QCAlgorithm):
 
         goog = self.add_equity("GOOG").symbol
 
-        option_chain = self.option_chain(goog)
+        option_chain = self.option_chain(goog, flatten=True)
 
         # Demonstration using data frame:
         df = option_chain.data_frame