Feat/type checking

.github/workflows/ci.yaml

@@ -24,7 +24,14 @@ jobs:
       - name: Install the project and dependencies
         run: uv sync --all-extras
 
-      - uses: astral-sh/ruff-action@v3
+      - name: Lint
+        uses: astral-sh/ruff-action@v3
+
+      - name: Type checkg
+        run: |
+          uv tool install ty
+          ty check src
+          ty check tests
 
       - name: Run tests with coverage
         run: uv run coverage run -m pytest

src/ConditionalGMM/UnivariateGMM.py

@@ -1,7 +1,8 @@
 """Helpful functions that can only be computed (easily) for univarate GMMs."""
 
 import numpy as np
-import scipy as sp
+from scipy.optimize import newton
+from scipy.stats import norm
 
 
 class UniGMM:
@@ -49,7 +50,7 @@ def pdf(self, x):
         assert np.ndim(x) < 2
         # TODO vectorize
         pdfs = np.array(
-            [sp.stats.norm.pdf(x, mi, vi) for mi, vi in zip(self.means, self.variances)]
+            [norm.pdf(x, mi, vi) for mi, vi in zip(self.means, self.variances)]
         )
         return np.dot(self.weights, pdfs)
 
@@ -76,7 +77,7 @@ def cdf(self, x):
 
         """
         cdfs = np.array(
-            [sp.stats.norm.cdf(x, mi, vi) for mi, vi in zip(self.means, self.variances)]
+            [norm.cdf(x, mi, vi) for mi, vi in zip(self.means, self.variances)]
         )
         return np.dot(self.weights, cdfs)
 
@@ -103,9 +104,7 @@ def ppf(self, q):
             (float or array-like) quantile corresponding to `q`
 
         """
-        return sp.optimize.newton(
-            func=lambda x: self.cdf(x) - q, x0=self.mean(), fprime=self.pdf
-        )
+        return newton(func=lambda x: self.cdf(x) - q, x0=self.mean(), fprime=self.pdf)
 
     def mean(self):
         """Mean of the RV for the GMM.

src/ConditionalGMM/condGMM.py

@@ -1,7 +1,7 @@
 """Conditional Gaussian mixture model."""
 
 import numpy as np
-import scipy as sp
+from scipy.stats import multivariate_normal
 
 from ConditionalGMM.MNorm import CondMNorm
 
@@ -103,7 +103,7 @@ def unconditional_pdf_x2(self, x2=None, component_probs=False):
 
         probs = w * np.array(
             [
-                sp.stats.multivariate_normal.pdf(x2, mean=mus[i], cov=covs[i])
+                multivariate_normal.pdf(x2, mean=mus[i], cov=covs[i])
                 for i in range(len(w))
             ]
         )

tests/test_UGMM.py

@@ -2,7 +2,7 @@
 
 import numpy as np
 import numpy.testing as npt
-import scipy as sp
+from scipy.stats import norm
 
 from ConditionalGMM import UnivariateGMM
 
@@ -25,7 +25,7 @@ def test_uggm_pdf():
         pdf = ugmm.pdf(x)
         truepdf = np.dot(
             np.array(weights),
-            np.array([sp.stats.norm.pdf(x, mi, vi) for mi, vi in zip(means, vars)]),
+            np.array([norm.pdf(x, mi, vi) for mi, vi in zip(means, vars)]),
         )
         npt.assert_equal(pdf, truepdf)