File size: 4,592 Bytes
9c31a35
3d7c303
9bfcbfa
dca02e2
3d7c303
 
 
7d4300a
9c31a35
51a6b05
 
 
9c31a35
7d4300a
9c31a35
51a6b05
9c31a35
 
 
7d4300a
9c31a35
7d4300a
c7187a6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9bfcbfa
b07eb2d
7d4300a
 
dca02e2
7d4300a
 
 
 
9bfcbfa
7d4300a
 
 
9bfcbfa
dca02e2
 
 
7d4300a
 
 
c7187a6
 
dca02e2
9bfcbfa
dca02e2
 
9bfcbfa
7d4300a
c7187a6
7d4300a
9bfcbfa
962c25c
 
 
932dcf5
962c25c
 
 
 
 
 
 
 
 
 
 
 
 
d4d95e5
 
 
 
 
 
c3a1736
d4d95e5
 
 
 
 
 
 
 
 
dca02e2
 
 
d4d95e5
 
 
 
c7187a6
dca02e2
c3a1736
d398bf9
 
c3a1736
 
d4d95e5
 
c7187a6
d4d95e5
 
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
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
import unittest
import numpy as np
import pandas as pd
from pysr import sympy2torch, PySRRegressor
import torch
import sympy


class TestTorch(unittest.TestCase):
    def setUp(self):
        np.random.seed(0)

    def test_sympy2torch(self):
        x, y, z = sympy.symbols("x y z")
        cosx = 1.0 * sympy.cos(x) + y
        X = torch.tensor(np.random.randn(1000, 3))
        true = 1.0 * torch.cos(X[:, 0]) + X[:, 1]
        torch_module = sympy2torch(cosx, [x, y, z])
        self.assertTrue(
            np.all(np.isclose(torch_module(X).detach().numpy(), true.detach().numpy()))
        )

    def test_pipeline_pandas(self):
        X = pd.DataFrame(np.random.randn(100, 10))
        equations = pd.DataFrame(
            {
                "Equation": ["1.0", "cos(x1)", "square(cos(x1))"],
                "MSE": [1.0, 0.1, 1e-5],
                "Complexity": [1, 2, 3],
            }
        )

        equations["Complexity MSE Equation".split(" ")].to_csv(
            "equation_file.csv.bkup", sep="|"
        )

        model = PySRRegressor(
            model_selection="accuracy",
            equation_file="equation_file.csv",
            extra_sympy_mappings={},
            output_torch_format=True,
        )
        # Because a model hasn't been fit via the `fit` method, some
        # attributes will not/cannot be set. For the purpose of
        # testing, these attributes will be set manually here.
        model.fit(X, y=np.ones(X.shape[0]), from_equation_file=True)
        model.refresh()

        tformat = model.pytorch()
        self.assertEqual(str(tformat), "_SingleSymPyModule(expression=cos(x1)**2)")
        np.testing.assert_almost_equal(
            tformat(torch.tensor(X.values)).detach().numpy(),
            np.square(np.cos(X.values[:, 1])),  # Selection 1st feature
            decimal=4,
        )

    def test_pipeline(self):
        X = np.random.randn(100, 10)
        equations = pd.DataFrame(
            {
                "Equation": ["1.0", "cos(x1)", "square(cos(x1))"],
                "MSE": [1.0, 0.1, 1e-5],
                "Complexity": [1, 2, 3],
            }
        )

        equations["Complexity MSE Equation".split(" ")].to_csv(
            "equation_file.csv.bkup", sep="|"
        )

        model = PySRRegressor(
            model_selection="accuracy",
            equation_file="equation_file.csv",
            extra_sympy_mappings={},
            output_torch_format=True,
        )

        model.fit(X, y=np.ones(X.shape[0]), from_equation_file=True)
        model.refresh()

        tformat = model.pytorch()
        self.assertEqual(str(tformat), "_SingleSymPyModule(expression=cos(x1)**2)")
        np.testing.assert_almost_equal(
            tformat(torch.tensor(X)).detach().numpy(),
            np.square(np.cos(X[:, 1])),  # 2nd feature
            decimal=4,
        )

    def test_mod_mapping(self):
        x, y, z = sympy.symbols("x y z")
        expression = x**2 + sympy.atanh(sympy.Mod(y + 1, 2) - 1) * 3.2 * z

        module = sympy2torch(expression, [x, y, z])

        X = torch.rand(100, 3).float() * 10

        true_out = (
            X[:, 0] ** 2 + torch.atanh(torch.fmod(X[:, 1] + 1, 2) - 1) * 3.2 * X[:, 2]
        )
        torch_out = module(X)

        np.testing.assert_array_almost_equal(
            true_out.detach(), torch_out.detach(), decimal=4
        )

    def test_custom_operator(self):
        X = np.random.randn(100, 3)

        equations = pd.DataFrame(
            {
                "Equation": ["1.0", "mycustomoperator(x1)"],
                "MSE": [1.0, 0.1],
                "Complexity": [1, 2],
            }
        )

        equations["Complexity MSE Equation".split(" ")].to_csv(
            "equation_file_custom_operator.csv.bkup", sep="|"
        )

        model = PySRRegressor(
            model_selection="accuracy",
            equation_file="equation_file_custom_operator.csv",
            extra_sympy_mappings={"mycustomoperator": sympy.sin},
            extra_torch_mappings={"mycustomoperator": torch.sin},
            output_torch_format=True,
        )
        model.fit(X, y=np.ones(X.shape[0]), from_equation_file=True)
        model.refresh()
        self.assertEqual(str(model.sympy()), "sin(x1)")
        # Will automatically use the set global state from get_hof.

        tformat = model.pytorch()
        self.assertEqual(str(tformat), "_SingleSymPyModule(expression=sin(x1))")
        np.testing.assert_almost_equal(
            tformat(torch.tensor(X)).detach().numpy(),
            np.sin(X[:, 1]),
            decimal=4,
        )