[M] Mode experiment scripts

experiments/action_sizes.py

@@ -0,0 +1,69 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from pathlib import Path
+
+import numpy as np
+import matplotlib.pyplot as plt
+
+@dataclass
+class Statistics:
+    mean: float
+    median: float
+    lower_quartile: float
+    upper_quartile: float
+    iqr: float
+    minimum: float
+    maximum: float
+    count: int
+    total: float
+    stddev: float
+
+    def get_metric_6(self) -> tuple[float, float, float, float, float, float]:
+        return self.mean, self.median, self.minimum, self.maximum, self.lower_quartile, self.upper_quartile
+
+    def print(self, dec: int = 2):
+        print(f'> Mean: {round(self.mean, dec)}, Median: {round(self.median, dec)}')
+        print(f'> Min: {round(self.minimum, dec)}, Max: {round(self.maximum, dec)}')
+        print(f'> Q1: {round(self.lower_quartile, dec)}, Q3: {round(self.upper_quartile, dec)}')
+        print(f'> StdDev: {round(self.stddev, dec)}, IQR: {round(self.iqr, dec)}')
+        print(f'> N: {self.count}')
+
+
+def _calc_col_stats_helper(col: np.ndarray) -> tuple[float, float, float, float, float, float, float, int, float, float]:
+    q1 = np.quantile(col, 0.25)
+    q3 = np.quantile(col, 0.75)
+    return (
+        float(np.mean(col)),
+        float(np.median(col)),
+        float(q1),
+        float(q3),
+        float(q3 - q1),
+        float(np.min(col)),
+        float(np.max(col)),
+        len(col),
+        float(np.sum(col)),
+        float(np.std(col))
+    )
+
+
+def calc_col_stats(col: np.ndarray | list) -> Statistics:
+    """
+    Compute statistics for a data column
+    :param col: Input column (tested on 1D array)
+    :return: Statistics
+    """
+    if isinstance(col, list):
+        col = np.array(col)
+    return Statistics(*_calc_col_stats_helper(col))
+
+
+if __name__ == '__main__':
+    txt = Path('action-sizes.log').read_text('utf-8').split('\n')
+    nums = [int(line) for line in txt if line.isnumeric()]
+    # print(nums)
+
+    calc_col_stats(nums).print()
+    plt.hist(nums, bins=50)
+    plt.show()
+

experiments/diff-absolute.csv

@@ -0,0 +1,40 @@
+,Date,Number of Patches
+0,2017-03-01,3149
+1,2017-06-01,3150
+2,2017-09-01,3146
+3,2017-12-01,3139
+4,2018-03-01,3146
+5,2018-04-01,3146
+6,2018-06-01,3148
+7,2018-07-01,3153
+8,2018-09-01,3148
+9,2018-10-01,3142
+10,2018-12-01,3142
+11,2019-01-01,3149
+12,2019-03-01,3147
+13,2019-04-01,3151
+14,2019-06-01,3148
+15,2019-07-01,3147
+16,2019-09-01,3152
+17,2019-10-01,3148
+18,2019-12-01,3146
+19,2020-01-01,3148
+20,2020-03-01,3146
+21,2020-04-01,3147
+22,2020-06-01,3146
+23,2020-07-01,3149
+24,2020-09-01,3140
+25,2020-10-01,3152
+26,2020-12-01,3136
+27,2021-01-01,3149
+28,2021-03-01,3146
+29,2021-04-01,3146
+30,2021-06-01,3148
+31,2021-07-01,3150
+32,2021-09-01,3138
+33,2021-10-01,3146
+34,2021-12-01,3148
+35,2022-01-01,3148
+36,2022-03-01,3142
+37,2022-04-01,3152
+38,2022-06-01,3151