Giray18 / sports_app_data_analysis_tool

A repo that has a python class structure that facilitates analysis of semi-structured (JSON) football dataset.

App Data (JSON) Analysis Tool

Motivation

Motivation for creating this repo is to create a python class that has methods which can bring insights from semi-structured dataset gathered from an sports app.

Scenario (Current Situation)

A BI team who is looking for to gather quick insights from a semi structured data applying to a data analyst for a tool creation to gather needed insight quickly. Used dataset covering 19 weeks of 2022-2023 Premier league season

Source Dataset Definition (RAW)

Source data located on (https://raw.githubusercontent.com/Giray18/sports_app_data_analysis_tool/main/epl_2022_2023_07_02_2023.json)

Source Dataset Definition Flattened Dataframe

• Columns of Flattened Dataframe
  • Index Col: Game Id
  • event: All actions taken in a game stored in a list
  • matchweek : matchweek of season
  • team1_name : home team name
  • team1_startings : starting lineup of home team saved in a list
  • team1_subs : subs of home team saved in a list
  • team1_stats : home team general stats on game e.g possesion, shots on target etc.
  • team2_name : away team name
  • team2_startings : starting lineup of away team saved in a list
  • team2_subs : subs of away team saved in a list
  • team2_stats : away team general stats on game e.g possesion, shots on target etc.

Needed Insights

• A method that brings team name as parameter and returns amount of goals that selected team scored
• Percentage of total goals in league of selected team
• Total shoots team made during entire season
• Shot to Goal accuracy of team
• Top 10 players that has most starting lineups
• Score distribution of all teams in the league in a dataset

Task Executed (Target Situation)

• For to get needed insights from source data 2 python class and 1 python module created and used.
• Two class called goal analyzer and lineups located in stats finder module (https://github.com/Giray18/sports_app_data_analysis_tool/tree/main/stats_finder) created and used.
• 1 function (read_df_json) from python module (dat) located on https://github.com/Giray18/sports_app_data_analysis_tool/tree/main/dat used to read JSON data into flattened pandas dataframe structure.
• Usage examples of classes can be found on jupyter notebook located on (https://github.com/Giray18/sports_app_data_analysis_tool/blob/main/playground.ipynb)
• As a result quick data analysis of semi-structured data done by a python class and methods within class.