mlbend-metrics

Search uses release 0.1.0.0 · 42 definitions

Releases (1)

0.1.0.0 indexed

Definitions (42)

main.bend

type Bad
def same_of1 when a and b are the same label, 0 otherwise
def same
def countthe number of samples
def count_true.gothe number of samples whose true label is k
def count_true
def count_pred.gothe number of samples whose predicted label is k
def count_pred
def count_both.gothe number of samples whose true label is k and whose predicted label is k too: the true positives of the class k
def count_both
def count_right.gothe number of samples predicted right
def count_right
law count_lengthLAW 1 (count_length): the number of samples is the length of the list of labels.
type Sum
def sum_add
def sq_sum.gothe sum of the squared differences of the pairs walked
def abs_sum.gothe sum of the absolute differences of the pairs walked
def ratio_ofthe ratio a over b, or 0.0 when b is zero
def ratio
def meanthe mean of a sum over n pairs
def same_len_ofthe number of pairs to walk, or why the two sequences do not agree
def same_len
def not_empty_ofthe number of pairs to walk, or why there are none
def not_emptythe number of pairs, or why there is no answer
def pairsthe number of pairs the two sequences hold, or why there is no answer
def mse.gothe mean squared error between the targets and the predictions
def mse
def mae.gothe mean absolute error between the targets and the predictions
def mae
def rmse.gothe root mean squared error between the targets and the predictions
def rmse
def accuracy.gothe fraction of the samples predicted right
def accuracy
def precision.gothe fraction of the samples predicted as k that are really k: the precision of the class k
def precision
def recall.gothe fraction of the samples really k that are predicted as k: the recall of the class k
def recall
def f1_num_ofthe harmonic mean of two scores
def f1_num
def f1_ofthe harmonic mean of a precision and a recall
def f1.gothe harmonic mean of the precision and the recall of the class k
def f1