// Determine our alpha parameter as |U| / (|U| + |D|). Note that because we
// append to X_training and remove from X_uncertainty within `ranked_batch`,
// :alpha: is not fixed throughout our model"s lifetime.
n_labeled, n_unlabeled = X_training.shape[0], X_uncertainty.shape[0]
alpha = n_unlabeled / (n_unlabeled + n_labeled)
// Isolate our original unlabeled records from their predicted class uncertainty.
unlabeled_records, uncertainty_scores = X_uncertainty[:, :-1], X_uncertainty[:, -1]
After Change
// we selected that row for labeling.
n_labeled_records, _ = X_training.shape
X_uncertainty = X_uncertainty[~np.isnan(X_uncertainty).all(axis=1)]
n_unlabeled, _ = X_uncertainty.shape
// Determine our alpha parameter as |U| / (|U| + |D|). Note that because we
// append to X_training and remove from X_uncertainty within `ranked_batch`,
