some pep8 fixes .... but probably more to come

Author: yamins81

scikits/learn/svm/src/liblinear/_liblinear.c

@@ -20,7 +20,7 @@ cdef extern from "linear.h":
     void destroy_param (parameter *)
 
 cdef extern from "liblinear_helper.c":
-    void copy_w(char *, model *, int,int,int)
+    void copy_w(char *, model *, int)
     parameter *set_parameter (int, double, double, int,
                              char *, char *)
     problem *set_problem (char *, char *, np.npy_intp *, double)
@@ -89,11 +89,11 @@ def train_wrap ( np.ndarray[np.float64_t, ndim=2, mode='c'] X,
     if bias > 0: nr_feature = nr_feature + 1
     if nr_class == 2:
         w = np.empty((1, nr_feature),order='F')
-        copy_w(w.data, model, nr_feature,0,0)
+        copy_w(w.data, model, nr_feature)
     else:
         len_w = (nr_class) * nr_feature
         w = np.empty((nr_class, nr_feature),order='F')
-        copy_w(w.data, model, len_w,nr_class,nr_feature)
+        copy_w(w.data, model, len_w)
 
     cdef np.ndarray[np.int32_t, ndim=1, mode='c'] label
     label = np.empty(nr_class, dtype=np.int32)
@@ -146,11 +146,11 @@ def csr_train_wrap ( int n_features,
     if bias > 0: nr_feature = nr_feature + 1
     if nr_class == 2:
         w = np.empty((1, nr_feature),order='F')
-        copy_w(w.data, model, nr_feature,0,0)
+        copy_w(w.data, model, nr_feature)
     else:
         len_w = (nr_class * nr_feature)
         w = np.empty((nr_class, nr_feature),order='F')
-        copy_w(w.data, model, len_w,nr_class,nr_feature)
+        copy_w(w.data, model, len_w)
 
     cdef np.ndarray[np.int32_t, ndim=1, mode='c'] label
     label = np.empty((nr_class), dtype=np.int32)

scikits/learn/svm/src/liblinear/_liblinear.pyx

@@ -210,7 +210,7 @@ void cfort(double *data, int len, int l1, int l2)
 
 }
 
-void copy_w(char *data, struct model *model, int len, int l1, int l2)
+void copy_w(char *data, struct model *model, int len)
 {
     memcpy(data, model->w, len * sizeof(double));