Repository navigation
Expand file tree
/
Copy pathscoring_algorithm.py
More file actions
119 lines (95 loc) · 3.6 KB
/
Copy pathscoring_algorithm.py
File metadata and controls
119 lines (95 loc) · 3.6 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
"""
| developed by: markmelnic
| original repo: https://github.com/markmelnic/Scoring-Algorithm
Analyse data using a range based percentual proximity algorithm
and calculate the linear maximum likelihood estimation.
The basic principle is that all values supplied will be broken
down to a range from ``0`` to ``1`` and each column's score will be added
up to get the total score.
Example for data of vehicles
::
price|mileage|registration_year
20k |60k |2012
22k |50k |2011
23k |90k |2015
16k |210k |2010
We want the vehicle with the lowest price,
lowest mileage but newest registration year.
Thus the weights for each column are as follows:
``[0, 0, 1]``
"""
def get_data(source_data: list[list[float]]) -> list[list[float]]:
"""
>>> get_data([[20, 60, 2012],[23, 90, 2015],[22, 50, 2011]])
[[20.0, 23.0, 22.0], [60.0, 90.0, 50.0], [2012.0, 2015.0, 2011.0]]
"""
data_lists: list[list[float]] = []
for data in source_data:
for i, el in enumerate(data):
if len(data_lists) < i + 1:
data_lists.append([])
data_lists[i].append(float(el))
return data_lists
def calculate_each_score(
data_lists: list[list[float]], weights: list[int]
) -> list[list[float]]:
"""
>>> calculate_each_score([[20, 23, 22], [60, 90, 50], [2012, 2015, 2011]],
... [0, 0, 1])
[[1.0, 0.0, 0.33333333333333337], [0.75, 0.0, 1.0], [0.25, 1.0, 0.0]]
"""
score_lists: list[list[float]] = []
for dlist, weight in zip(data_lists, weights):
mind = min(dlist)
maxd = max(dlist)
score: list[float] = []
# for weight 0 score is 1 - actual score
if weight == 0:
for item in dlist:
try:
score.append(1 - ((item - mind) / (maxd - mind)))
except ZeroDivisionError:
score.append(1)
elif weight == 1:
for item in dlist:
try:
score.append((item - mind) / (maxd - mind))
except ZeroDivisionError:
score.append(0)
# weight not 0 or 1
else:
msg = f"Invalid weight of {weight:f} provided"
raise ValueError(msg)
score_lists.append(score)
return score_lists
def generate_final_scores(score_lists: list[list[float]]) -> list[float]:
"""
>>> generate_final_scores([[1.0, 0.0, 0.33333333333333337],
... [0.75, 0.0, 1.0],
... [0.25, 1.0, 0.0]])
[2.0, 1.0, 1.3333333333333335]
"""
# initialize final scores
final_scores: list[float] = [0 for i in range(len(score_lists[0]))]
for slist in score_lists:
for j, ele in enumerate(slist):
final_scores[j] = final_scores[j] + ele
return final_scores
def procentual_proximity(
source_data: list[list[float]], weights: list[int]
) -> list[list[float]]:
"""
| `weights` - ``int`` list
| possible values - ``0`` / ``1``
* ``0`` if lower values have higher weight in the data set
* ``1`` if higher values have higher weight in the data set
>>> procentual_proximity([[20, 60, 2012],[23, 90, 2015],[22, 50, 2011]], [0, 0, 1])
[[20, 60, 2012, 2.0], [23, 90, 2015, 1.0], [22, 50, 2011, 1.3333333333333335]]
"""
data_lists = get_data(source_data)
score_lists = calculate_each_score(data_lists, weights)
final_scores = generate_final_scores(score_lists)
# append scores to source data
for i, ele in enumerate(final_scores):
source_data[i].append(ele)
return source_data