patch hmmlearn, fix dependencies

MACS3/Signal/HMMR_HMM.pyx

@@ -1,6 +1,6 @@
 # cython: language_level=3
 # cython: profile=True
-# Time-stamp: <2022-10-04 15:14:15 Tao Liu>
+# Time-stamp: <2024-02-18 16:21:00 Tao Liu>
 
 """Module description:
 
@@ -20,7 +20,8 @@ from math import sqrt
 import numpy as np
 cimport numpy as np
 from cpython cimport bool
-from hmmlearn import hmm
+from hmmlearn import hmm, _utils
+from sklearn import cluster
 import json
 # from hmmlearn cimport hmm
 
@@ -49,6 +50,35 @@ cdef inline float get_weighted_density( int x, float m, float v, w ):
 # ------------------------------------
 # Classes
 # ------------------------------------
+
+class GaussianHMM_modified( hmm.GaussianHMM ):
+    def _init(self, X, lengths=None):
+        super()._init(X, lengths)
+        # we will overwrite initial means_ and covars_
+        kmeans = cluster.KMeans(n_clusters=self.n_components,
+                                random_state=self.random_state,
+                                n_init=10)  # https://github.com/hmmlearn/hmmlearn/pull/545
+                                # the idea is to do the random seeds
+                                # for 10 times orginally, hmmlearn 0.3
+                                # will do this only once.  However,
+                                # due to the change in scikit-learn
+                                # 1.3, the random seeding in KMeans
+                                # will generate different results with
+                                # previous scikit-learn. It will make
+                                # the results irreproducible between
+                                # sklearn <1.3 and sklearn
+                                # >=1.3. Hopefully, if we choose to do
+                                # the process 10 times, the results
+                                # will be more similar.
+        kmeans.fit(X)
+        self.means_ = kmeans.cluster_centers_
+
+        cv = np.cov(X.T) + self.min_covar * np.eye(X.shape[1])
+        if not cv.shape:
+            cv.shape = (1, 1)
+        self.covars_ = \
+            _utils.distribute_covar_matrix_to_match_covariance_type( cv, self.covariance_type, self.n_components ).copy()
+
 # ------------------------------------
 # public functions
 # ------------------------------------
@@ -60,7 +90,7 @@ cpdef hmm_training( list training_data, list training_data_lengths, int n_states
     # according to base documentation, if init_prob not stated, it is set to be equally likely for any state (1/ # of components)
     # if we have other known parameters, we should set these (ie: means_weights, covariance_type etc.)
     rs = np.random.RandomState(np.random.MT19937(np.random.SeedSequence(random_seed)))
-    hmm_model = hmm.GaussianHMM( n_components= n_states, covariance_type = covar, random_state = rs, verbose = False )
+    hmm_model = GaussianHMM_modified( n_components= n_states, covariance_type = covar, random_state = rs, verbose = False )
     hmm_model = hmm_model.fit( training_data, training_data_lengths )
     assert hmm_model.n_features == 4
     return hmm_model
@@ -91,7 +121,7 @@ cpdef void hmm_model_save( str model_file, object hmm_model, int hmm_binsize, in
 cpdef list hmm_model_init( str model_file ):
     with open( model_file ) as f:
         m = json.load( f )
-        hmm_model = hmm.GaussianHMM( n_components=3, covariance_type=m["covariance_type"] )
+        hmm_model = GaussianHMM_modified( n_components=3, covariance_type=m["covariance_type"] )
         hmm_model.startprob_ = np.array(m["startprob"])
         hmm_model.transmat_ = np.array(m["transmat"])
         hmm_model.means_ = np.array(m["means"])

pyproject.toml

@@ -1,3 +1,3 @@
 [build-system]
-requires=['setuptools>=60.0', 'numpy>=1.24', 'scipy>=1.10', 'cykhash>=2.0,<3.0', 'Cython~=3.0', 'scikit-learn>=1.2,<1.4', 'hmmlearn>=0.3']
+requires=['setuptools>=60.0', 'numpy>=1.24.2', 'scipy>=1.11.4', 'cykhash>=2.0,<3.0', 'Cython~=3.0', 'scikit-learn>=1.2.1', 'hmmlearn>=0.3']
 

requirements.txt

@@ -1,7 +1,7 @@
 Cython~=3.0
-numpy>=1.24
-scipy>=1.10
-scikit-learn>=1.2,<1.4
+numpy>=1.24.2
+scipy>=1.11.4
+scikit-learn>=1.2.1
 hmmlearn>=0.3
 cykhash>=2.0,<3.0 
 pytest>=7.0

setup.py

@@ -37,10 +37,10 @@
               'Programming Language :: Python :: 3.12',
               'Programming Language :: Cython', ]
 
-install_requires = [ "numpy>=1.24",
-                     "scipy>=1.10",
+install_requires = [ "numpy>=1.24.2",
+                     "scipy>=1.11.4",
                      "hmmlearn>=0.3",
-                     "scikit-learn>=1.2,<1.4",
+                     "scikit-learn>=1.2.1",
                      "cykhash>=2.0,<3.0"]