8782146110

Committed 22 Apr 2024 10:00AM UTC coverage: 86.975%. First build

Build # 8782146110

Build Type

Pull #600

github

Committed by

mikkokotila

Commit Message

change the mentioned stats in readme

Pull Request Pull Request #600: few small changes to README.md

Run Details

1429 of 1643 relevant lines covered (86.98%)

5.2 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

75.0

/talos/utils/validation_split.py

def validation_split(self):

    '''Defines the attributes `x_train`, `y_train`, `x_val` and `y_val`.
    The validation sets are determined by the attribute val_split,
    which is a number in (0, 1) which determines the proportion of
    the input data to be allocated for cross-validation.'''

    # data input is list but multi_input is not set to True
    if isinstance(self.x, list) and self.multi_input is False:

        raise TypeError("For multi-input x, set multi_input to True")

    # If split is done in `Scan()` do nothing
    if self.custom_val_split:

        self.x_train = self.x
        self.y_train = self.y

        return self

    # Otherwise start by shuffling
    import wrangle
    self.x, self.y = wrangle.array_random_shuffle(x=self.x,
                                                  y=self.y,
                                                  multi_input=self.multi_input)

    # deduce the midway point for input data
    limit = int(len(self.y) * (1 - self.val_split))

    # handle the case where x is multi-input
    if self.multi_input:

        self.x_train = []
        self.x_val = []

        for ar in self.x:
            self.x_train.append(ar[:limit])
            self.x_val.append(ar[limit:])

    # handle the case where x is not multi-input
    else:

        self.x_train = self.x[:limit]
        self.x_val = self.x[limit:]

    # handle y data same for both cases
    self.y_train = self.y[:limit]
    self.y_val = self.y[limit:]

    return self


def kfold(x, y, folds=10, shuffled=True, multi_input=False):

    import wrangle

    # data input is list but multi_input is not set to True
    if isinstance(x, list) and multi_input is False:
        raise TypeError("For multi-input x, set multi_input to True")

    if shuffled is True:
        x, y = wrangle.array_random_shuffle(x, y, multi_input)

    out_x = []
    out_y = []

    # establish the fold size
    y_len = len(y)
    step = int(y_len / folds)

    lo = 0
    hi = step

    # create folds one by one
    for _i in range(folds):

        # handle the case for multi-input model
        if multi_input:
            fold_x = []
            for ar in x:
                fold_x.append(ar[lo:hi])
            out_x.append(fold_x)

        # handle the case where model is not multi-input
        else:
            out_x.append(x[lo:hi])

        out_y.append(y[lo:hi])

        lo += step
        hi += step

    return out_x, out_y

1	def validation_split(self):	6✔
2
3	'''Defines the attributes `x_train`, `y_train`, `x_val` and `y_val`.
4	The validation sets are determined by the attribute val_split,
5	which is a number in (0, 1) which determines the proportion of
6	the input data to be allocated for cross-validation.'''
7
8	# data input is list but multi_input is not set to True
9	if isinstance(self.x, list) and self.multi_input is False:	6✔
10
11	raise TypeError("For multi-input x, set multi_input to True")	×
12
13	# If split is done in `Scan()` do nothing
14	if self.custom_val_split:	6✔
15
16	self.x_train = self.x	6✔
17	self.y_train = self.y	6✔
18
19	return self	6✔
20
21	# Otherwise start by shuffling
22	import wrangle	6✔
23	self.x, self.y = wrangle.array_random_shuffle(x=self.x,	6✔
24	y=self.y,
25	multi_input=self.multi_input)
26
27	# deduce the midway point for input data
28	limit = int(len(self.y) * (1 - self.val_split))	6✔
29
30	# handle the case where x is multi-input
31	if self.multi_input:	6✔
32
33	self.x_train = []	×
34	self.x_val = []	×
35
36	for ar in self.x:	×
37	self.x_train.append(ar[:limit])	×
38	self.x_val.append(ar[limit:])	×
39
40	# handle the case where x is not multi-input
41	else:
42
43	self.x_train = self.x[:limit]	6✔
44	self.x_val = self.x[limit:]	6✔
45
46	# handle y data same for both cases
47	self.y_train = self.y[:limit]	6✔
48	self.y_val = self.y[limit:]	6✔
49
50	return self	6✔
51
52
53	def kfold(x, y, folds=10, shuffled=True, multi_input=False):	6✔
54
55	import wrangle	6✔
56
57	# data input is list but multi_input is not set to True
58	if isinstance(x, list) and multi_input is False:	6✔
59	raise TypeError("For multi-input x, set multi_input to True")	×
60
61	if shuffled is True:	6✔
62	x, y = wrangle.array_random_shuffle(x, y, multi_input)	6✔
63
64	out_x = []	6✔
65	out_y = []	6✔
66
67	# establish the fold size
68	y_len = len(y)	6✔
69	step = int(y_len / folds)	6✔
70
71	lo = 0	6✔
72	hi = step	6✔
73
74	# create folds one by one
75	for _i in range(folds):	6✔
76
77	# handle the case for multi-input model
78	if multi_input:	6✔
79	fold_x = []	×
80	for ar in x:	×
81	fold_x.append(ar[lo:hi])	×
82	out_x.append(fold_x)	×
83
84	# handle the case where model is not multi-input
85	else:
86	out_x.append(x[lo:hi])	6✔
87
88	out_y.append(y[lo:hi])	6✔
89
90	lo += step	6✔
91	hi += step	6✔
92
93	return out_x, out_y	6✔

autonomio / talos / 8782146110

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous