Shot dists.py

import logging
import pandas as pd
import numpy as np
import utils
from nba_api.stats.static import players
from nba_api.stats.static import teams
import plotly.express as px

logger = logging.getLogger(__name__)
root_logger = logging.getLogger()
root_logger.setLevel(logging.INFO)
sh = logging.StreamHandler()
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
sh.setFormatter(formatter)
root_logger.addHandler(sh)

desired_width = 320
pd.set_option('display.max_columns', 20)
pd.set_option('display.width', desired_width)

df = utils.load_pbp_jsons()
df = utils.add_pbp_oncourt_columns(df)

# Only filter for shot data
shots_df = df[(df["actionType"] == "2pt") | (df["actionType"] == "3pt")]
shots_df["shot_made"] = False
shots_df.loc[shots_df["shotResult"] == "Made", "shot_made"] = True

# Group shots by distance
# Put twos and threes into separate bins
zone_bins = {"2pt": [0, 4, 12], "3pt": [22, 23.75, 26]}

shots_df = shots_df.assign(shot_zone=np.nan)
counter = 1
for zone_bin_key, zone_bin_vals in zone_bins.items():
    for i in range(len(zone_bin_vals)):
        if i == len(zone_bin_vals)-1:
            zone_name = f"{zone_bin_vals[i]}+"
            max_dist = 94
        else:
            zone_name = f"{zone_bin_vals[i]}-{zone_bin_vals[i + 1]}"
            max_dist = zone_bin_vals[i + 1]

        shots_df.loc[
            (shots_df["actionType"] == zone_bin_key) &
            (shots_df["shotDistance"] >= zone_bin_vals[i]) &
            (shots_df["shotDistance"] < max_dist),
            "shot_zone"
        ] = f"{counter}_{zone_bin_key}_{zone_name}"
        counter += 1