Replace GETC to SAMPLE_WEIGHT

Author: MechCoder

sklearn/svm/liblinear.pxd

@@ -20,10 +20,10 @@ cdef extern from "src/liblinear/linear.h":
 cdef extern from "src/liblinear/liblinear_helper.c":
  void copy_w(void *, model *, int)
  parameter *set_parameter(int, double, double, int, char *, char *, int, int, double)
- problem *set_problem (char *, char *, np.npy_intp *, double, char *)
+ problem *set_problem (char *, char *, np.npy_intp *, double, double *)
  problem *csr_set_problem (char *values, np.npy_intp *n_indices,
  char *indices, np.npy_intp *n_indptr, char *indptr, char *Y,
- np.npy_intp n_features, double bias, char *)
+ np.npy_intp n_features, double bias, double *)
 
  model *set_model(parameter *, char *, np.npy_intp *, char *, double)
 

sklearn/svm/liblinear.pyx

@@ -11,11 +11,11 @@ cimport liblinear
 np.import_array()
 
 
-def train_wrap(X, np.ndarray[np.float64_t,  ndim=1, mode='c'] Y,
+def train_wrap(X, np.ndarray[np.float64_t, ndim=1, mode='c'] Y,
  bint is_sparse, int solver_type, double eps, double bias,
  double C, np.ndarray[np.float64_t, ndim=1] class_weight,
  int max_iter, unsigned random_seed, double epsilon,
- np.ndarray[np.float64_t, ndim=1] sample_weight):
+ np.ndarray[np.float64_t, ndim=1, mode='c'] sample_weight):
  cdef parameter *param
  cdef problem *problem
  cdef model *model

sklearn/svm/src/liblinear/liblinear_helper.c

@@ -124,7 +124,7 @@ static struct feature_node **csr_to_sparse(double *values,
  return sparse;
 }
 
-struct problem * set_problem(char *X,char *Y, npy_intp *dims, double bias, char* sample_weight)
+struct problem * set_problem(char *X,char *Y, npy_intp *dims, double bias, double* sample_weight)
 {
  struct problem *problem;
  /* not performant but simple */
@@ -141,7 +141,7 @@ struct problem * set_problem(char *X,char *Y, npy_intp *dims, double bias, char*
  problem->y = (double *) Y;
  problem->x = dense_to_sparse((double *) X, dims, bias);
  problem->bias = bias;
- problem->sample_weight = (double *) sample_weight;
+ problem->sample_weight = sample_weight;
  if (problem->x == NULL) { 
  free(problem);
  return NULL;
@@ -152,7 +152,7 @@ struct problem * set_problem(char *X,char *Y, npy_intp *dims, double bias, char*
 
 struct problem * csr_set_problem (char *values, npy_intp *n_indices,
 char *indices, npy_intp *n_indptr, char *indptr, char *Y,
- npy_intp n_features, double bias, char *sample_weight) {
+ npy_intp n_features, double bias, double *sample_weight) {
 
  struct problem *problem;
  problem = malloc (sizeof (struct problem));
@@ -169,7 +169,7 @@ struct problem * csr_set_problem (char *values, npy_intp *n_indices,
  problem->x = csr_to_sparse((double *) values, n_indices, (int *) indices,
 n_indptr, (int *) indptr, bias, n_features);
  problem->bias = bias;
- problem->sample_weight = (double *) sample_weight;
+ problem->sample_weight = sample_weight;
 
  if (problem->x == NULL) {
  free(problem);

sklearn/svm/src/liblinear/linear.cpp

@@ -1244,8 +1244,9 @@ static int solve_l2r_l1l2_svr(
 // See Algorithm 5 of Yu et al., MLJ 2010
 
 
-#define GETC(i) upper_bound[y[i]+1]*sample_weight[i]
-// To support weights for instances (sample_weight*class_weight), use GETC(i) (i)
+#define SAMPLE_WEIGHT(i) upper_bound[y[i]+1]*sample_weight[i]
+// To support weights for instances, use SAMPLE_WEIGHT(i)
+// Each instance is weighted by sample_weight*class_weight)
 
 int solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, double Cn,
  int max_iter)
@@ -1276,12 +1277,12 @@ int solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, dou
 }
 
 // Initial alpha can be set here. Note that
-// 0 < alpha[i] < GETC(i)
-// alpha[2*i] + alpha[2*i+1] = GETC(i)
+// 0 < alpha[i] < SAMPLE_WEIGHT(i)
+// alpha[2*i] + alpha[2*i+1] = SAMPLE_WEIGHT(i)
 for(i=0; i<l; i++)
 {
-alpha[2*i] = min(0.001*GETC(i), 1e-8);
-alpha[2*i+1] = GETC(i) - alpha[2*i];
+alpha[2*i] = min(0.001*SAMPLE_WEIGHT(i), 1e-8);
+alpha[2*i+1] = SAMPLE_WEIGHT(i) - alpha[2*i];
 }
 
 for(i=0; i<w_size; i++)
@@ -1313,7 +1314,7 @@ int solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, dou
 {
 i = index[s];
 schar yi = y[i];
-double C = GETC(i);
+double C = SAMPLE_WEIGHT(i);
 double ywTx = 0, xisq = xTx[i];
 feature_node *xi = prob->x[i];
 while (xi->index != -1)
@@ -1396,7 +1397,7 @@ int solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, dou
 v *= 0.5;
 for(i=0; i<l; i++)
 v += alpha[2*i] * log(alpha[2*i]) + alpha[2*i+1] * log(alpha[2*i+1])
-- GETC(i) * log(GETC(i));
+- SAMPLE_WEIGHT(i) * log(SAMPLE_WEIGHT(i));
 info("Objective value = %lf\n", v);
 
 delete [] xTx;
@@ -1705,9 +1706,10 @@ static int solve_l1r_l2_svc(
 // solution will be put in w
 //
 // See Yuan et al. (2011) and appendix of LIBLINEAR paper, Fan et al. (2008)
-#undef GETC
-#define GETC(i) C[y[i]+1]*sample_weight[i]
-// To support weights (sample_weight*class_weight) for instances, use GETC(i) (i)
+#undef SAMPLE_WEIGHT
+#define SAMPLE_WEIGHT(i) C[y[i]+1]*sample_weight[i]
+// To support weights for instances, use SAMPLE_WEIGHT(i)
+// Each instance is weighted by (class_weight*sample_weight)
 
 static int solve_l1r_lr(
 const problem *prob_col, double *w, double eps,
@@ -1777,16 +1779,16 @@ static int solve_l1r_lr(
 double val = x->value;
 exp_wTx[ind] += w[j]*val;
 if(y[ind] == -1)
-xjneg_sum[j] += GETC(ind)*val;
+xjneg_sum[j] += SAMPLE_WEIGHT(ind)*val;
 x++;
 }
 }
 for(j=0; j<l; j++)
 {
 exp_wTx[j] = exp(exp_wTx[j]);
 double tau_tmp = 1/(1+exp_wTx[j]);
-tau[j] = GETC(j)*tau_tmp;
-D[j] = GETC(j)*exp_wTx[j]*tau_tmp*tau_tmp;
+tau[j] = SAMPLE_WEIGHT(j)*tau_tmp;
+D[j] = SAMPLE_WEIGHT(j)*exp_wTx[j]*tau_tmp*tau_tmp;
 }
 
 while(newton_iter < max_newton_iter)
@@ -1962,7 +1964,7 @@ static int solve_l1r_lr(
 negsum_xTd = 0;
 for(int i=0; i<l; i++)
 if(y[i] == -1)
-negsum_xTd += GETC(i)*xTd[i];
+negsum_xTd += SAMPLE_WEIGHT(i)*xTd[i];
 
 int num_linesearch;
 for(num_linesearch=0; num_linesearch < max_num_linesearch; num_linesearch++)
@@ -1973,7 +1975,7 @@ static int solve_l1r_lr(
 {
 double exp_xTd = exp(xTd[i]);
 exp_wTx_new[i] = exp_wTx[i]*exp_xTd;
-cond += GETC(i)*log((1+exp_wTx_new[i])/(exp_xTd+exp_wTx_new[i]));
+cond += SAMPLE_WEIGHT(i)*log((1+exp_wTx_new[i])/(exp_xTd+exp_wTx_new[i]));
 }
 
 if(cond <= 0)
@@ -1985,8 +1987,8 @@ static int solve_l1r_lr(
 {
 exp_wTx[i] = exp_wTx_new[i];
 double tau_tmp = 1/(1+exp_wTx[i]);
-tau[i] = GETC(i)*tau_tmp;
-D[i] = GETC(i)*exp_wTx[i]*tau_tmp*tau_tmp;
+tau[i] = SAMPLE_WEIGHT(i)*tau_tmp;
+D[i] = SAMPLE_WEIGHT(i)*exp_wTx[i]*tau_tmp*tau_tmp;
 }
 break;
 }
@@ -2053,9 +2055,9 @@ static int solve_l1r_lr(
 }
 for(j=0; j<l; j++)
 if(y[j] == 1)
-v += GETC(j)*log(1+1/exp_wTx[j]);
+v += SAMPLE_WEIGHT(j)*log(1+1/exp_wTx[j]);
 else
-v += GETC(j)*log(1+exp_wTx[j]);
+v += SAMPLE_WEIGHT(j)*log(1+exp_wTx[j]);
 
 info("Objective value = %lf\n", v);
 info("#nonzeros/#features = %d/%d\n", nnz, w_size);
@@ -2496,6 +2498,7 @@ model* train(const problem *prob, const parameter *param)
 free(sub_prob.x);
 free(sub_prob.y);
 free(weighted_C);
+delete[] sample_weight;
 }
 return model_;
 }