diff options
| author | Gertjan van den Burg <burg@ese.eur.nl> | 2014-07-02 17:09:38 +0200 |
|---|---|---|
| committer | Gertjan van den Burg <burg@ese.eur.nl> | 2014-07-02 17:09:38 +0200 |
| commit | 75a91a963c979b0d657b8839036cd82f6ae5f24b (patch) | |
| tree | 1fd14e14d6cf0e877c07ecc95258e3968460cc0a /src/msvmmaj_train_dataset.c | |
| parent | set element of category matrix to zero explicitly, not calloced (diff) | |
| download | gensvm-75a91a963c979b0d657b8839036cd82f6ae5f24b.tar.gz gensvm-75a91a963c979b0d657b8839036cd82f6ae5f24b.zip | |
seed V based on scale of X
Diffstat (limited to 'src/msvmmaj_train_dataset.c')
| -rw-r--r-- | src/msvmmaj_train_dataset.c | 12 |
1 files changed, 6 insertions, 6 deletions
diff --git a/src/msvmmaj_train_dataset.c b/src/msvmmaj_train_dataset.c index 8d23684..7fa7316 100644 --- a/src/msvmmaj_train_dataset.c +++ b/src/msvmmaj_train_dataset.c @@ -335,7 +335,7 @@ void consistency_repeats(struct Queue *q, long repeats, TrainType traintype) model->m = task->train_data->m; model->K = task->train_data->K; msvmmaj_allocate_model(model); - msvmmaj_seed_model_V(NULL, model); + msvmmaj_seed_model_V(NULL, model, task->train_data); } time[i] = 0.0; @@ -356,7 +356,7 @@ void consistency_repeats(struct Queue *q, long repeats, TrainType traintype) note("%3.3f\t", p); // this is done because if we reuse the V it's not a // consistency check - msvmmaj_seed_model_V(NULL, model); + msvmmaj_seed_model_V(NULL, model, task->train_data); } for (r=0; r<repeats; r++) { std[i] += pow(matrix_get( @@ -512,7 +512,7 @@ void start_training_cv(struct Queue *q) model->m = task->train_data->m; model->K = task->train_data->K; msvmmaj_allocate_model(model); - msvmmaj_seed_model_V(NULL, model); + msvmmaj_seed_model_V(NULL, model, task->train_data); main_s = clock(); while (task) { @@ -576,7 +576,7 @@ void start_training_tt(struct Queue *q) seed_model->m = task->train_data->m; seed_model->K = task->train_data->K; msvmmaj_allocate_model(seed_model); - msvmmaj_seed_model_V(NULL, seed_model); + msvmmaj_seed_model_V(NULL, seed_model, task->train_data); main_s = clock(); while (task) { @@ -594,7 +594,7 @@ void start_training_tt(struct Queue *q) msvmmaj_allocate_model(model); msvmmaj_initialize_weights(task->train_data, model); - msvmmaj_seed_model_V(seed_model, model); + msvmmaj_seed_model_V(seed_model, model, task->train_data); fid = MSVMMAJ_OUTPUT_FILE; MSVMMAJ_OUTPUT_FILE = NULL; @@ -607,7 +607,7 @@ void start_training_tt(struct Queue *q) if (task->test_data->y != NULL) total_perf = msvmmaj_prediction_perf(task->test_data, predy); - msvmmaj_seed_model_V(model, seed_model); + msvmmaj_seed_model_V(model, seed_model, task->train_data); msvmmaj_free_model(model); free(predy); |