lsapy 0.3.0__tar.gz → 0.3.1__tar.gz

{lsapy-0.3.0 → lsapy-0.3.1}/CHANGELOG.rst

@@ -2,7 +2,15 @@
 Changelog
 =========
 
-v0.3.0 (2025-10-11)
+v0.3.1 (2025-11-14)
+-------------------
+Contributor to this version: Baptiste Hamon (@baptistehamon).
+
+Bug fixes
+^^^^^^^^^
+* The `np.vectorize` decorator has been removed of standardization functions to fix fitting functions issues (issue `#101 <https://github.com/baptistehamon/lsapy/issues/101>`_, PR `#102 <https://github.com/baptistehamon/lsapy/pull/102>`_).
+
+v0.3.0 (2025-11-10)
 -------------------
 Contributor to this version: Baptiste Hamon (@baptistehamon).
 

{lsapy-0.3.0 → lsapy-0.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lsapy
-Version: 0.3.0
+Version: 0.3.1
 Summary: A Python package for Land Suitability Analysis.
 Keywords: LSAPy,Land Suitability Analysis,Geospatial Analysis,GIS,Geography
 Author-email: Baptiste Hamon <baptiste.hamon@pg.canterbury.ac.nz>

{lsapy-0.3.0 → lsapy-0.3.1}/pyproject.toml

@@ -83,7 +83,7 @@ Issues = "https://github.com/baptistehamon/lsapy/issues"
 Zenodo = "https://doi.org/10.5281/zenodo.15015110"
 
 [tool.bumpversion]
-current_version = "0.3.0"
+current_version = "0.3.1"
 commit = true
 commit_args = "--no-verify --signoff"
 tag = false

{lsapy-0.3.0 → lsapy-0.3.1}/src/lsapy/__init__.py

@@ -6,4 +6,4 @@ from lsapy.lsa import LandSuitabilityAnalysis
 
 __author__ = "Baptiste Hamon"
 __email__ = "baptiste.hamon@pg.canterbury.ac.nz"
-__version__ = "0.3.0"
+__version__ = "0.3.1"

{lsapy-0.3.0 → lsapy-0.3.1}/src/lsapy/functions/membership.py

@@ -281,14 +281,14 @@ def fit_membership(x, y=None, fit_on: str | list[str] = "all", plot: bool = Fals
     --------
     >>> from lsapy.functions.membership import fit_membership
     >>> fit_membership([1, 3, 5, 7, 10])  # doctest: +ELLIPSIS
-    (<numpy.vectorize object at 0x...>, array([1., 5.]))
+    (<function vetharaniam2022_eq5 at 0x...>, array([-2.789...,  4.864...]))
 
     By default, the function will fit all available membership equations. If you want to fit only specific equations,
     you can specify it using the `fit_on` parameter: "all", "sigmoid_like", "gaussian_like", or a list of equations.
     The default `y` values can also be changed.
 
     >>> fit_membership(x=[1, 3, 5, 5, 7, 9], y=[0, 0.5, 1, 1, 0.5, 0], fit_on="gaussian_like")  # doctest: +ELLIPSIS
-    (<numpy.vectorize object at 0x...>, array([0.667..., 4.941..., 0.846...]))
+    (<function vetharaniam2024_eq10 at 0x...>, array([0.382..., 4.972..., 0.939...]))
     """
     if y is None:
         y = [0, 0.25, 0.5, 0.75, 1]

{lsapy-0.3.0 → lsapy-0.3.1}/src/lsapy/standardize.py

@@ -68,7 +68,6 @@ def boolean(x, op: str, thresh: int | float, skipna: bool = True):
 
 
 @declare_equation("categorical")
-@np.vectorize(otypes=[np.float32])
 def discrete(x, rules: dict[str | int, int | float]) -> np.ndarray:
     """
     Discrete function.
@@ -88,12 +87,11 @@ def discrete(x, rules: dict[str | int, int | float]) -> np.ndarray:
     np.ndarray
         Mapped output values.
     """
-    return rules.get(x, np.nan)
+    return np.vectorize(rules.get)(x, np.nan)
 
 
 @declare_equation("sigmoid")
-@np.vectorize(otypes=[np.float32])
-def logistic(x, a, b):
+def logistic(x, a, b) -> np.ndarray:
     r"""
     Logistic function capped to 1.
 
@@ -122,11 +120,11 @@ def logistic(x, a, b):
 
         f(x) = \frac{1}{1 + e^{-a(x - b)}}
     """
-    return 1 / (1 + np.exp(-a * (x - b)))
+    return 1 / (1 + np.exp(-a * np.subtract(x, b)))
 
 
 @declare_equation("sigmoid")
-def sigmoid(x):
+def sigmoid(x) -> np.ndarray:
     r"""
     Logistic sigmoid function.
 
@@ -159,8 +157,7 @@ def sigmoid(x):
 
 
 @declare_equation("sigmoid", "VTR22_eq3")
-@np.vectorize(otypes=[np.float32])
-def vetharaniam2022_eq3(x, a, b):
+def vetharaniam2022_eq3(x, a, b) -> np.ndarray:
     r"""
     Sigmoid like function.
 
@@ -200,12 +197,11 @@ def vetharaniam2022_eq3(x, a, b):
     ----------
     :cite:cts:`vetharaniam_lsa_2022`
     """
-    return np.exp(a * (x - b)) / (1 + np.exp(a * (x - b)))
+    return np.exp(a * np.subtract(x, b)) / (1 + np.exp(a * np.subtract(x, b)))
 
 
 @declare_equation("sigmoid", "VTR22_eq5")
-@np.vectorize(otypes=[np.float32])
-def vetharaniam2022_eq5(x, a, b):
+def vetharaniam2022_eq5(x, a, b) -> np.ndarray:
     r"""
     Sigmoid like function.
 
@@ -243,8 +239,7 @@ def vetharaniam2022_eq5(x, a, b):
 
 
 @declare_equation("gaussian", "VTR24_eq8")
-@np.vectorize(otypes=[np.float32])
-def vetharaniam2024_eq8(x, a, b, c):
+def vetharaniam2024_eq8(x, a, b, c) -> np.ndarray:
     r"""
     Gaussian like function.
 
@@ -280,12 +275,11 @@ def vetharaniam2024_eq8(x, a, b, c):
     ----------
     :cite:cts:`vetharaniam_lsa_2024`
     """
-    return np.exp(-a * np.power(x - b, c))
+    return np.exp(-a * np.power(np.subtract(x, b), c))
 
 
 @declare_equation("gaussian", "VTR24_eq10")
-@np.vectorize(otypes=[np.float32])
-def vetharaniam2024_eq10(x, a, b, c):
+def vetharaniam2024_eq10(x, a, b, c) -> np.ndarray:
     r"""
     Gaussian like function.
 

{lsapy-0.3.0 → lsapy-0.3.1}/tests/test_fitting.py

@@ -97,7 +97,7 @@ class TestFit:
         x = [-10, -6.5, -5, -3.5, 0]
         f, p = std.fit(x, kind="all", plot=True)
         assert f == std.logistic
-        np.testing.assert_array_almost_equal(p, [0.91, -5.0], decimal=2)
+        np.testing.assert_array_almost_equal(p, [0.76, -5.0], decimal=2)
 
     def test_skipped(self):
         with pytest.warns(UserWarning, match="No methods to fit. Skipping: vetharaniam2024_eq8, vetharaniam2024_eq10."):

{lsapy-0.3.0 → lsapy-0.3.1}/tests/test_membership_fitting.py

@@ -98,7 +98,7 @@ class TestFitMembership:
         x = [-10, -6.5, -5, -3.5, 0]
         f, p = mbs.fit_membership(x, fit_on="all", plot=True)
         assert f == std.logistic
-        np.testing.assert_array_almost_equal(p, [0.91, -5.0], decimal=2)
+        np.testing.assert_array_almost_equal(p, [0.76, -5.0], decimal=2)
 
     def test_skipped(self):
         with pytest.warns(UserWarning, match="No methods to fit. Skipping: vetharaniam2024_eq8, vetharaniam2024_eq10."):