Create test_string_to_function.py

tests/test_string_to_function.py

@@ -0,0 +1,138 @@
+import numpy as np
+import pytest
+
+from innate.regression.methods import generate_specific_function  
+from innate.regression.methods import extract_coef_names  
+from innate.regression.methods import extract_variables_names   
+from innate.regression.methods import create_coef_dict
+
+
+
+
+expression = 'a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)'
+coef_names = extract_coef_names(expression)
+print(coef_names)
+coef_values = [1,2,3]
+print(coef_values)
+#X['a', 'b', 'c']
+# Extract variables ending with '_range' (default suffix)
+range_variables = extract_variables_names(expression)
+print(range_variables) 
+#['variable1_range', 'variable2_range']
+
+
+print(create_coef_dict(coef_names, coef_values))
+#{'a': 1, 'b': 2, 'c': 3}
+
+
+
+expression = 'a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)'
+
+# Fixed values for a, b, c
+coefficients = {'a': 1, 'b': 4, 'c': 7}
+
+# Specify the variable names that should vary
+variable_names = ['variable1_range', 'variable2_range']  # These names should match usage in expression
+
+# Create a function that varies with specified variables
+@pytest.fixture 
+def specific_function():
+    expression = 'a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)'
+
+    # Fixed values for a, b, c
+    coefficients = {'a': 1, 'b': 4, 'c': 7}
+
+    # Specify the variable names that should vary
+    variable_names = ['variable1_range', 'variable2_range']
+    return generate_specific_function(expression, coefficients, variable_names) 
+#print( str(specific_function([10000, 20000, 30000],[10000, 20000, 30000])))
+@pytest.mark.parametrize(
+    "test, expected",
+    [
+        ([10000, 20000, 30000], [5.0, 5.107209969647869, 5.67318211637097])
+    ]
+
+
+
+)
+def test_specific_function(test, expected):
+    specific_function = generate_specific_function(expression, coefficients, variable_names) 
+    np.testing.assert_almost_equal(
+        specific_function([10000, 20000, 30000],[10000, 20000, 30000]), np.array(expected), decimal=7
+    )
+
+
+expression = 'a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)'
+
+# Fixed values for a, b, c
+coefficients = {'a': 1, 'b': 4, 'c': 7}
+
+# Specify the variable names that should vary
+variable_names = ['variable1', 'variable2']  # These names should match usage in expression
+
+# Create a function that varies with specified variables
+specific_function = generate_specific_function(expression, coefficients, variable_names)
+
+# Example usage with arrays for each variable
+variable1_values = np.array([10000, 20000, 30000])  # Example values for variable1
+variable2_values = np.array([10000, 20000, 30000])  # Example values for variable2
+
+
+
+expression = 'a + b / (variable1/10000.0) + c * np.log10(variable2/10000)'
+
+# Extract variables
+@pytest.mark.parametrize(
+    "test, expected",
+    [
+        ('a + b / (variable1/10000.0) + c * np.log10(variable2/10000)', ['a', 'b', 'c'])
+    ]
+
+)
+def test_extract_coef_names(test, expected):
+    np.testing.assert_equal(
+        extract_coef_names( test ),  expected) 
+
+
+
+
+@pytest.mark.parametrize(
+    "test, expected",
+    [
+        ((coef_names,coef_values),{'a': 1, 'b': 2, 'c': 3})
+    ]
+
+)
+def test_create_coef_dict(test, expected):
+    expression = 'a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)'
+    coef_names = extract_coef_names(expression)
+    coef_values = [1,2,3]
+    np.testing.assert_equal(
+       create_coef_dict(coef_names,coef_values),expected
+    )
+
+@pytest.mark.parametrize(
+    "test, expected",
+    [
+        ('a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)',['variable1_range', 'variable2_range'])
+    ]
+
+)
+def test_extract_variables_names(test, expected):
+    np.testing.assert_equal(
+       extract_variables_names(test ), expected
+    )
+
+
+# Answer should be {'a': 1, 'b': 2, 'c': 3}
+
+# expression = 'a + b / (variable1_range/10000.0) + c * np.log10(variable2_range/10000)'
+
+# Extract variables ending with '_range' (default suffix)
+# range_variables = extract_variables_names(expression)
+# print(range_variables) 
+#['variable1_range', 'variable2_range']
+
+
+# create_coef_dict (coef_names, coef_values)
+#{'a': 1, 'b': 2, 'c': 3}