Lines Matching refs:dims
53 * inv_eigen_values - set of 1*dims matrices, <inv_eigen_values>[k] contains
173 m->cols != train_data.dims )
175 "floating-point matrix (CvMat) of 'nsamples' x 'dims' size" );
197 cov->rows != cov->cols || cov->cols != train_data.dims )
200 "floating-point matrix (CvMat) of 'dims' x 'dims'" );
219 int i, k, dims;
226 dims = means->cols;
229 CV_CALL( cvPreparePredictData( _sample, dims, 0, params.nclusters, _probs, &sample_data ));
232 size = sizeof(double) * (nclusters + dims);
241 diff = cvMat( 1, dims, CV_64FC1, (double*)buffer + nclusters );
254 for( i = 0; i < dims; i++ )
262 for( i = 0; i < dims; i++ )
266 for( i = 0; i < dims; i++ )
279 /* probability = (2*pi)^(-dims/2)*exp( -0.5 * cur ) */
317 int i, nsamples, nclusters, dims;
324 &train_data.count, &train_data.dims, &train_data.dims,
330 dims = train_data.dims;
343 CV_CALL( means = cvCreateMat( nclusters, dims, CV_64FC1 ));
346 params.cov_mat_type == COV_MAT_SPHERICAL ? 1 : dims, CV_64FC1 ));
352 CV_CALL( covs[i] = cvCreateMat( dims, dims, CV_64FC1 ));
353 CV_CALL( cov_rotate_mats[i] = cvCreateMat( dims, dims, CV_64FC1 ));
368 CvMat sample = cvMat( 1, dims, CV_32F );
406 int nclusters = params.nclusters, nsamples = train_data.count, dims = train_data.dims;
448 CV_CALL( tcov = cvCreateMat( dims, dims, CV_64FC1 ));
449 CV_CALL( w = cvCreateMat( dims, dims, CV_64FC1 ));
451 CV_CALL( u = cvCreateMat( dims, dims, CV_64FC1 ));
464 cvSetIdentity( covs[i], cvScalar(cvTrace(w).val[0]/dims) );
494 int nclusters = params.nclusters, nsamples = train_data.count, dims = train_data.dims;
499 CvMat src = cvMat( 1, dims, CV_32F );
500 CvMat dst = cvMat( 1, dims, CV_64F );
538 hdr[0] = cvMat( 1, dims, CV_32F );
585 int i, j, k, nsamples, dims;
592 dims = train_data.dims;
595 CV_CALL( centers = cvCreateMat( nclusters, dims, CV_64FC1 ));
596 CV_CALL( old_centers = cvCreateMat( nclusters, dims, CV_64FC1 ));
628 for( j = 0; j <= dims - 4; j += 4 )
638 for( ; j < dims; j++ )
669 for( j = 0; j <= dims - 4; j += 4 )
683 for( ; j < dims; j++ )
697 for( j = 0; j < dims; j++ )
710 for( j = 0; j < dims; j++ )
718 for( j = 0; j < dims; j++ )
789 int nsamples = train_data.count, dims = train_data.dims, nclusters = params.nclusters;
791 double min_det_value = MAX( DBL_MIN, pow( min_variation, dims ));
792 double likelihood_bias = -CV_LOG2PI * (double)nsamples * (double)dims / 2., _log_likelihood = -DBL_MAX;
809 d = cvTrace(*covs).val[0]/dims;
812 log_weight = pow( d, dims*0.5 );
824 for( j = 0, det = 1.; j < dims; j++ )
848 CV_CALL( covs_item = cvCreateMat( dims, dims, CV_64FC1 ));
849 CV_CALL( centered_sample = cvCreateMat( 1, dims, CV_64FC1 ));
851 CV_CALL( samples = cvCreateMat( nsamples, dims, CV_64FC1 ));
861 for( j = 0; j < dims; j++ )
877 for( j = 0, det = 1.; j < dims; j++ )
890 d = cvTrace(covs[k]).val[0]/(double)dims;
905 cvScale( log_det, log_det, dims );
931 for( j = 0; j < dims; j++ )
935 for( j = 0; j < dims; j++ )
1010 for( j = 0; j < dims; j++ )
1019 d = w_data[0]/(double)dims;
1029 for( j = 0, det = 1.; j < dims; j++ )
1040 cvScale( log_det, log_det, dims );
