Spaces:
Sleeping
Sleeping
| #!/usr/bin/env python3 | |
| # -*- coding: utf-8 -*- | |
| """ | |
| Created on 2024-07-19 20:32:30 Friday | |
| @author: Nikhil Kapila | |
| """ | |
| import numpy as np | |
| import pandas as pd | |
| def sliding_windows(data:pd.DataFrame, lookback:int=30)->tuple[np.array, np.array, pd.Index]: | |
| X, y, timestamps = [], [], [] | |
| for i in range(len(data) - lookback): | |
| X.append(data.iloc[i:i + lookback].values) | |
| y.append(data.iloc[i + lookback]) | |
| timestamps.append(data.index[i + lookback]) | |
| return np.array(X), np.array(y), pd.Index(timestamps) | |
| def resampler(df, time='h'): | |
| times = ['h', 'm', 'd'] | |
| if time not in times: raise ValueError | |
| return df.resample(time).sum() | |
| def df_from_np(values, timestamps, value_col='predicted'): | |
| if len(values.shape) == 2: values1 = values.flatten() | |
| return pd.DataFrame({ | |
| 'timestamp': timestamps, | |
| f'{value_col}': values | |
| }) | |