Deduplication logic for City Marshal Evictions Dataset

[sarahJune1]

The deduplication logic for evictions dataset must take into account both the court index number and borough code. The current logic assumes that court index number is unique across the dataset and therefore is dropping data.

The scope of this issue would add the borough code and court index number to the logic.

Additional context:

• There are about 100 rows per month that are missing in the nycdb but are present in the Open Data source file due to the current deduplication logic that uses only the court index number.
• This dataset is used in the displacement project

Link to EDA notebook: https://colab.research.google.com/drive/1sLET77zixEa_bDzbaqsWuUwbsUKwhM7z?usp=sharing