WCFA_BA_Team1

BA Team1 Readme

Our Tier Intervals

• "challenger_grandmaster": Includes challengers and grandmasters
• "master_diamond": Includes masters and upper diamonds
• "diamond_emerald": Includes lower diamonds and upper emeralds
• "emerald_platinum": Includes lower emeralds and upper platinums
• "gold": Include golds
• "silver": Include silvers

Raw Data (Riot API Data) Collecting & Preprocessing Module

1. get_match_ids.py

   • Get match IDs from the starting point user who were in that tier.

2. check_matchid_duplication.py

   • Eliminate duplicate match data within different tier data (usually in continuous tier) to prevent data duplication.

3. get_match_data.py

   • Get match data from match IDs collected before.

4. process_raw_data.py

   • Process one game data to obtain difference values among blue team and red team, generating two JSON files (blue team data & red team).

5. get_csv_format.py

   • Combine JSON files collected before and get an integrated CSV file for each tier interval.

6. Check_tier.py

   • Verify that collected data for each tier interval is truly data for that tier.

Other Python Files

• get_match_from_challenger_gm.py

  • Set aside for collecting upper echelons' data, specifically challenger and grandmaster tier's data.

• concatenate_csv’s.py

  • Concatenate each tier interval's CSV file to create one integrated CSV file.

• Csv_functions.py

  • Contains CSV functions.

• json_functions.py

  • Contains JSON functions.

• riot_api_functions.csv

  • Contains Riot API functions.

• combine_result_csv.py

  • Concatenate various result CSV files then extract one integrated file.

• Divide_csv_by_duration.csv

  • Divide processed raw data according to the game duration (Early, Mid, Late).

• model.py

  • Conduct machine learning based on game data, conducting logistic regression with a pipeline + hyperparameter grid, then outputting the CSV format results.

Data Python Files

• Factor_list.csv.py

  • List up and classify the used factors in our analysis.

• Champ_tier_final.py

  • List the champion tier according to tier interval.

Directories

1. "All_data_ver1", "matched_and_puuid_ver1"

   • Deprecated directories and files. Wrongly collected data is included.

2. Raw JSON File Directories

   • Contain each tier interval's match data JSON files from Riot API.
   • A. "Challenger_grandmaster_data"
   • B. "Master_diamond_data"
   • C. "Diamond_emerald_data"
   • D. "Emerald_platinum_data"
   • E. "Gold_data"
   • F. "Silver_data"

3. "check_lists"

   • Contains CSV-format data, and each row indicates the average tier of 10 users who participated in the game.

4. "matchid_and_puuid_ver2"

   • Contains used match IDs and user's PUUIDs for collecting the entire dataset from Riot API.

5. "processed_data"

   • Contains JSON files for each tier interval. Each JSON file is blue or red team's data, output from "process_raw_data.py".

6. "processed_raw_data"

   • Contains CSV files for each tier interval. Each CSV file is generated by integrating the JSON files in "processed_data", output from "get_csv_format.py".

7. "processed_time_duration_data"

   • Contains CSV files for each tier interval and game duration interval (early, mid, and late). Each CSV file is generated by dividing the CSV files in "processed_raw_data" by "divide_csv_by_duration.py".

8. Result CSV Directories

   • A. "results": Contains model's result files using CSV files in "processed_raw_data".
   • B. "time_duration_results_csv": Contains model's result files using CSV files in "processed_time_duration_data".