do v0.12.3

CHANGES.md

@@ -1,11 +1,11 @@
-# version 0.12.2
+# version 0.12.3
 
 - add prediction intervals to `LSBoostRegressor` (split conformal prediction, 
   split conformal prediction with Kernel Density Estimation, and split 
   conformal prediction bootstrap)
   see `examples/lsboost_regressor_pi.py` for examples 
 - do not rescale columns with zero variance in `LSBoostRegressor` and `LSBoostClassifier`
-- faster ridge regression for `LSBoostRegressor`
+- faster ridge regression for `LSBoostRegressor` and `LSBoostClassifier`
 
 # version 0.9.0
 

mlsauce/lasso/_lasso.py

@@ -6,6 +6,7 @@
 from sklearn.base import RegressorMixin
 from numpy.linalg import inv
 from . import _lassoc as mo
+from ..utils import get_beta
 
 if platform.system() in ("Linux", "Darwin"):
     import jax.numpy as jnp
@@ -83,7 +84,8 @@ def fit(self, X, y, **kwargs):
         Xy2 = 2 * Xy
 
         if self.backend == "cpu":                               
-            beta0, _, _, _ = np.linalg.lstsq(X_, centered_y, rcond=None)            
+            #beta0, _, _, _ = np.linalg.lstsq(X_, centered_y, rcond=None)            
+            beta0 = get_beta(X_, centered_y)            
             if len(np.asarray(y).shape) == 1:
                 res = mo.get_beta_1D(
                     beta0=np.asarray(beta0),

mlsauce/ridge/_ridge.py

@@ -5,6 +5,7 @@
 from sklearn.base import RegressorMixin
 from numpy.linalg import inv
 from . import _ridgec as mo
+from ..utils import get_beta  
 
 if platform.system() in ("Linux", "Darwin"):
     import jax.numpy as jnp
@@ -73,13 +74,15 @@ def fit(self, X, y, **kwargs):
                 eye_term = np.sqrt(self.reg_lambda) * np.eye(X.shape[1])
                 X_ = np.row_stack((X_, eye_term))
                 y_ = np.concatenate((centered_y, np.zeros(X.shape[1])))
-                self.beta, _, _, _ = np.linalg.lstsq(X_, y_, rcond=None)            
+                #self.beta, _, _, _ = np.linalg.lstsq(X_, y_, rcond=None)            
+                self.beta = get_beta(X_, y_)            
             else: 
                 try: 
                     eye_term = np.sqrt(self.reg_lambda) * np.eye(X.shape[1])
                     X_ = np.row_stack((X_, eye_term))
                     y_ = np.row_stack((centered_y, np.zeros((eye_term.shape[0], centered_y.shape[1]))))
-                    self.beta, _, _, _ = np.linalg.lstsq(X_, y_, rcond=None)            
+                    #self.beta, _, _, _ = np.linalg.lstsq(X_, y_, rcond=None)            
+                    self.beta = get_beta(X_, y_)            
                 except Exception:
                     x = inv(mo.crossprod(X_) + self.reg_lambda * np.eye(X_.shape[1]))
                     hat_matrix = mo.tcrossprod(x, X_)

mlsauce/utils/__init__.py

@@ -1,6 +1,7 @@
 from .sampling.rowsubsampling import subsample
 from .misc.misc import merge_two_dicts, flatten, is_float, is_factor
 from .progress_bar import Progbar
+from .get_beta import get_beta
 
 __all__ = [
     "subsample",
@@ -9,4 +10,5 @@
     "is_float",
     "is_factor",
     "Progbar",
+    "get_beta"
 ]

mlsauce/utils/get_beta.py

@@ -0,0 +1,4 @@
+import numpy as np
+
+def get_beta(X, y):
+    return np.linalg.solve(X.T @ X, X.T @ y)

setup.py

@@ -38,7 +38,7 @@
 MAINTAINER_EMAIL = '[email protected]'
 LICENSE = 'BSD3 Clause Clear'
 
-__version__ = '0.12.2'
+__version__ = '0.12.3'
 
 VERSION = __version__