WIP: initial implementation of harmonised plots

bob/pad/base/script/__init__.py

+import numpy as np
+
+
+def remove_nan(scores):
+    """Removes the NaNs from the scores
+
+    Parameters
+    ----------
+    scores : numpy.array
+        The scores.
+
+    Returns
+    -------
+    scores : numpy.array
+        The scores excluding the NaNs.
+    sum_nans : int
+        The number of NaNs in the scores.
+    total : int
+        The total number of scores before removing the NaNs.
+    """
+    nans = np.isnan(scores)
+    return scores[np.where(~nans)], sum(nans), len(scores)
+
+
+def fta(neg_pos):
+    """Calculates the Failure To Acquire (FtA) rate.
+    Calculates how many NaNs are in neg and pos in total and returns the rate.
+
+    Parameters
+    ----------
+    neg_pos : (numpy.array, numpy.array)
+        A tuple of negative and positive scores
+
+    Returns
+    -------
+    neg : numpy.array
+        The negative scores with NaNs removed.
+    pos : numpy.array
+        The positive scores with NaNs removed.
+    fta : float
+        The FTA rate.
+    """
+    fta_sum, fta_total = 0, 0
+    neg, sum_nans, total = remove_nan(neg_pos[0])
+    fta_sum += sum_nans
+    fta_total += total
+    pos, sum_nans, total = remove_nan(neg_pos[1])
+    fta_sum += sum_nans
+    fta_total += total
+    return (neg, pos, fta_sum / fta_total)
+
+
+
 from . import spoof