Merge pull request #247 from jasmainak/check_distr

ENH add check for distr

Author: Pavan Ramkumar

pyglmnet/pyglmnet.py

@@ -9,6 +9,10 @@
 from .base import BaseEstimator, is_classifier, check_version
 
 
+ALLOWED_DISTRS = ['gaussian', 'binomial', 'softplus', 'poisson',
+                  'probit', 'gamma']
+
+
 def _lmb(distr, beta0, beta, X, eta):
     """Conditional intensity function."""
     z = beta0 + np.dot(X, beta)
@@ -333,7 +337,7 @@ class GLM(BaseEstimator):
     ----------
     distr : str
         distribution family can be one of the following
-        'gaussian' | 'binomial' | 'poisson' | 'softplus'
+        'gaussian' | 'binomial' | 'poisson' | 'softplus' | 'probit' | 'gamma'
         default: 'poisson'.
     alpha : float
         the weighting between L1 penalty and L2 penalty term
@@ -401,6 +405,10 @@ def __init__(self, distr='poisson', alpha=0.5,
         if not isinstance(max_iter, int):
             raise ValueError('max_iter must be of type int')
 
+        if distr not in ALLOWED_DISTRS:
+            raise ValueError('distr must be one of %s, Got '
+                             '%s' % (', '.join(ALLOWED_DISTRS), distr))
+
         self.distr = distr
         self.alpha = alpha
         self.reg_lambda = reg_lambda
@@ -845,7 +853,7 @@ class GLMCV(object):
     ----------
     distr : str
         distribution family can be one of the following
-        'gaussian' | 'binomial' | 'poisson' | 'softplus'
+        'gaussian' | 'binomial' | 'poisson' | 'softplus' | 'probit' | 'gamma'
         default: 'poisson'.
     alpha : float
         the weighting between L1 penalty and L2 penalty term
@@ -926,6 +934,10 @@ def __init__(self, distr='poisson', alpha=0.5,
         if not isinstance(reg_lambda, (list, np.ndarray)):
             reg_lambda = [reg_lambda]
 
+        if distr not in ALLOWED_DISTRS:
+            raise ValueError('distr must be one of %s, Got '
+                             '%s' % (', '.join(ALLOWED_DISTRS), distr))
+
         if not isinstance(max_iter, int):
             raise ValueError('max_iter must be of type int')
 

tests/test_pyglmnet.py

@@ -136,6 +136,9 @@ def test_group_lasso():
 
 def test_glmnet():
     """Test glmnet."""
+    assert_raises(ValueError, GLM, distr='blah')
+    assert_raises(ValueError, GLM, distr='gaussian', max_iter=1.8)
+
     n_samples, n_features = 100, 10
 
     # coefficients
@@ -192,6 +195,9 @@ def test_glmnet():
 
 def test_glmcv():
     """Test GLMCV class."""
+    assert_raises(ValueError, GLM, distr='blah')
+    assert_raises(ValueError, GLM, distr='gaussian', max_iter=1.8)
+
     scaler = StandardScaler()
     n_samples, n_features = 100, 10