Update app.py
Browse files
app.py
CHANGED
|
@@ -90,7 +90,8 @@ from datasets import load_dataset
|
|
| 90 |
### Import Datasets
|
| 91 |
season = 2025
|
| 92 |
level = 'mlb'
|
| 93 |
-
dataset = load_dataset('TJStatsApps/mlb_data', data_files=[f'{level}_pitch_data_{season}.csv' ])
|
|
|
|
| 94 |
dataset_train = dataset['train']
|
| 95 |
df = dataset_train.to_pandas().set_index(list(dataset_train.features.keys())[0]).reset_index(drop=True).drop_duplicates(subset=['play_id'],keep='last')
|
| 96 |
df['batter_name_team'] = df['batter_name'] + ' - ' + df['batter_team']
|
|
|
|
| 90 |
### Import Datasets
|
| 91 |
season = 2025
|
| 92 |
level = 'mlb'
|
| 93 |
+
# dataset = load_dataset('TJStatsApps/mlb_data', data_files=[f'{level}_pitch_data_{season}.csv' ])
|
| 94 |
+
dataset = pl.read_parquet(f"hf://datasets/TJStatsApps/mlb_data/data/{level}_pitch_data_{season}.parquet")
|
| 95 |
dataset_train = dataset['train']
|
| 96 |
df = dataset_train.to_pandas().set_index(list(dataset_train.features.keys())[0]).reset_index(drop=True).drop_duplicates(subset=['play_id'],keep='last')
|
| 97 |
df['batter_name_team'] = df['batter_name'] + ' - ' + df['batter_team']
|