efloch / dunder-mifflin Goto Github PK

In the root folder, run

python -m run

The script will be loaded, analyzed and key statistics will be saved into the data/processed/ folder.

• The data is first loaded and episode names are added to the raw table from a second dataset (process_script.py)
• Then each line is broken down into tokens, and words are counted (breakdown_line.py)
• Each line is processed to extract its sentiment score and the characters mentioned (analyze_line.py)
• The processed and anaylzed lines are saved processed_script.csv
• Finally, word and line counts are aggregated into multiple files and one master table is saved under all_counts.csv