@Override
public void modifyParmsForCrossValidationMainModel(ModelBuilder[] cvModelBuilders) {
_parms._overwrite_with_best_model = false;
if (_parms._stopping_rounds == 0 && _parms._max_runtime_secs == 0)
return; // No exciting changes to stopping conditions
// Extract stopping conditions from each CV model, and compute the best stopping answer
_parms._stopping_rounds = 0;
_parms._max_runtime_secs = 0;
double sum = 0;
for (ModelBuilder cvmb : cvModelBuilders)
sum += ((DeepLearningModel) DKV.getGet(cvmb.dest())).last_scored().epoch_counter;
_parms._epochs = sum / cvModelBuilders.length;
if (!_parms._quiet_mode) {
warn(
"_epochs",
"Setting optimal _epochs to "
+ _parms._epochs
+ " for cross-validation main model based on early stopping of cross-validation models.");
warn(
"_stopping_rounds",
"Disabling convergence-based early stopping for cross-validation main model.");
warn(
"_max_runtime_secs",
"Disabling maximum allowed runtime for cross-validation main model.");
}
}
/**
* Initialize the ModelBuilder, validating all arguments and preparing the training frame. This
* call is expected to be overridden in the subclasses and each subclass will start with
* "super.init();". This call is made by the front-end whenever the GUI is clicked, and needs to
* be fast; heavy-weight prep needs to wait for the trainModel() call.
*/
@Override
public void init(boolean expensive) {
super.init(expensive);
if ((_parms.start_column != null) && !_parms.start_column.isInt())
error("start_column", "start time must be null or of type integer");
if (!_parms.stop_column.isInt()) error("stop_column", "stop time must be of type integer");
if (!_parms.event_column.isInt() && !_parms.event_column.isCategorical())
error("event_column", "event must be of type integer or factor");
if (Double.isNaN(_parms.lre_min) || _parms.lre_min <= 0)
error("lre_min", "lre_min must be a positive number");
if (_parms.iter_max < 1) error("iter_max", "iter_max must be a positive integer");
final int MAX_TIME_BINS = 10000;
final long min_time =
(_parms.start_column == null)
? (long) _parms.stop_column.min()
: (long) _parms.start_column.min() + 1;
final int n_time = (int) (_parms.stop_column.max() - min_time + 1);
if (n_time < 1) error("start_column", "start times must be strictly less than stop times");
if (n_time > MAX_TIME_BINS)
error(
"stop_column",
"number of distinct stop times is "
+ n_time
+ "; maximum number allowed is "
+ MAX_TIME_BINS);
}
/**
* Initialize the ModelBuilder, validating all arguments and preparing the training frame. This
* call is expected to be overridden in the subclasses and each subclass will start with
* "super.init();". This call is made by the front-end whenever the GUI is clicked, and needs to
* be fast; heavy-weight prep needs to wait for the trainModel() call.
*
* <p>Validate the requested ntrees; precompute actual ntrees. Validate the number of classes to
* predict on; validate a checkpoint.
*/
@Override
public void init(boolean expensive) {
super.init(expensive);
if (H2O.ARGS.client && _parms._build_tree_one_node)
error("_build_tree_one_node", "Cannot run on a single node in client mode");
if (_vresponse != null) _vresponse_key = _vresponse._key;
if (_response != null) _response_key = _response._key;
if (_nclass > SharedTreeModel.SharedTreeParameters.MAX_SUPPORTED_LEVELS)
error("_nclass", "Too many levels in response column!");
if (_parms._min_rows < 0) error("_min_rows", "Requested min_rows must be greater than 0");
if (_parms._ntrees < 0 || _parms._ntrees > 100000)
error("_ntrees", "Requested ntrees must be between 1 and 100000");
_ntrees = _parms._ntrees; // Total trees in final model
if (_parms._checkpoint) { // Asking to continue from checkpoint?
Value cv = DKV.get(_parms._model_id);
if (cv != null) { // Look for prior model
M checkpointModel = cv.get();
if (_parms._ntrees < checkpointModel._output._ntrees + 1)
error(
"_ntrees",
"Requested ntrees must be between "
+ checkpointModel._output._ntrees
+ 1
+ " and 100000");
_ntrees = _parms._ntrees - checkpointModel._output._ntrees; // Needed trees
}
}
if (_parms._nbins <= 1) error("_nbins", "_nbins must be > 1.");
if (_parms._nbins >= 1 << 16) error("_nbins", "_nbins must be < " + (1 << 16));
if (_parms._nbins_cats <= 1) error("_nbins_cats", "_nbins_cats must be > 1.");
if (_parms._nbins_cats >= 1 << 16) error("_nbins_cats", "_nbins_cats must be < " + (1 << 16));
if (_parms._max_depth <= 0) error("_max_depth", "_max_depth must be > 0.");
if (_parms._min_rows <= 0) error("_min_rows", "_min_rows must be > 0.");
if (_parms._distribution == Distributions.Family.tweedie) {
_parms._distribution.tweedie.p = _parms._tweedie_power;
}
if (_train != null) {
double sumWeights =
_train.numRows() * (hasWeightCol() ? _train.vec(_parms._weights_column).mean() : 1);
if (sumWeights
< 2 * _parms._min_rows) // Need at least 2*min_rows weighted rows to split even once
error(
"_min_rows",
"The dataset size is too small to split for min_rows="
+ _parms._min_rows
+ ": must have at least "
+ 2 * _parms._min_rows
+ " (weighted) rows, but have only "
+ sumWeights
+ ".");
}
if (_train != null) _ncols = _train.numCols() - 1 - numSpecialCols();
}
/**
* Initialize the ModelBuilder, validating all arguments and preparing the training frame. This
* call is expected to be overridden in the subclasses and each subclass will start with
* "super.init();". This call is made by the front-end whenever the GUI is clicked, and needs to
* be fast; heavy-weight prep needs to wait for the trainModel() call.
*
* <p>Validate the requested ntrees; precompute actual ntrees. Validate the number of classes to
* predict on; validate a checkpoint.
*/
@Override
public void init(boolean expensive) {
super.init(expensive);
if (H2O.ARGS.client && _parms._build_tree_one_node)
error("_build_tree_one_node", "Cannot run on a single node in client mode");
if (_vresponse != null) _vresponse_key = _vresponse._key;
if (_response != null) _response_key = _response._key;
if (_parms._min_rows < 0) error("_min_rows", "Requested min_rows must be greater than 0");
if (_parms._ntrees < 0 || _parms._ntrees > MAX_NTREES)
error("_ntrees", "Requested ntrees must be between 1 and " + MAX_NTREES);
_ntrees = _parms._ntrees; // Total trees in final model
if (_parms.hasCheckpoint()) { // Asking to continue from checkpoint?
Value cv = DKV.get(_parms._checkpoint);
if (cv != null) { // Look for prior model
M checkpointModel = cv.get();
try {
_parms.validateWithCheckpoint(checkpointModel._parms);
} catch (H2OIllegalArgumentException e) {
error(e.values.get("argument").toString(), e.values.get("value").toString());
}
if (_parms._ntrees < checkpointModel._output._ntrees + 1)
error(
"_ntrees",
"If checkpoint is specified then requested ntrees must be higher than "
+ (checkpointModel._output._ntrees + 1));
// Compute number of trees to build for this checkpoint
_ntrees = _parms._ntrees - checkpointModel._output._ntrees; // Needed trees
}
}
if (_parms._nbins <= 1) error("_nbins", "_nbins must be > 1.");
if (_parms._nbins >= 1 << 16) error("_nbins", "_nbins must be < " + (1 << 16));
if (_parms._nbins_cats <= 1) error("_nbins_cats", "_nbins_cats must be > 1.");
if (_parms._nbins_cats >= 1 << 16) error("_nbins_cats", "_nbins_cats must be < " + (1 << 16));
if (_parms._max_depth <= 0) error("_max_depth", "_max_depth must be > 0.");
if (_parms._min_rows <= 0) error("_min_rows", "_min_rows must be > 0.");
if (_train != null) {
double sumWeights =
_train.numRows() * (hasWeightCol() ? _train.vec(_parms._weights_column).mean() : 1);
if (sumWeights
< 2 * _parms._min_rows) // Need at least 2*min_rows weighted rows to split even once
error(
"_min_rows",
"The dataset size is too small to split for min_rows="
+ _parms._min_rows
+ ": must have at least "
+ 2 * _parms._min_rows
+ " (weighted) rows, but have only "
+ sumWeights
+ ".");
}
if (_train != null) _ncols = _train.numCols() - 1 - numSpecialCols();
}
GridSearch(Key gkey, MP params, Map<String, Object[]> hyperSearch) {
super(
Key.<Grid>make("GridSearch_" + modelName() + Key.rand()),
gkey,
modelName() + " Grid Search");
_params = params;
_hyperSearch = hyper2doubles(hyperSearch);
// Count of models in this search
int work = 1;
for (double hparms[] : _hyperSearch) work *= hparms.length;
_total_models = work;
// Check all parameter combos for validity
double[] hypers = new double[_hyperSearch.length];
// FIXME: this expect finite space!
for (int[] hidx = new int[_hyperSearch.length]; hidx != null; hidx = nextModel(hidx)) {
ModelBuilder mb = getBuilder(params, hypers(hidx, hypers));
if (mb.error_count() > 0) throw new IllegalArgumentException(mb.validationErrors());
}
}
/**
* Initialize the ModelBuilder, validating all arguments and preparing the training frame. This
* call is expected to be overridden in the subclasses and each subclass will start with
* "super.init();". This call is made by the front-end whenever the GUI is clicked, and needs to
* be fast; heavy-weight prep needs to wait for the trainModel() call.
*
* <p>Validate the very large number of arguments in the DL Parameter directly.
*/
@Override
public void init(boolean expensive) {
super.init(expensive);
_parms.validate(this, expensive);
if (expensive && error_count() == 0) checkMemoryFootPrint();
}
/**
* @param hypers A set of hyper parameter values
* @return A Future of a model run with these parameters, typically built on demand and not cached
* - expected to be an expensive operation. If the model in question is "in progress", a 2nd
* build will NOT be kicked off. This is a non-blocking call.
*/
private ModelBuilder startBuildModel(MP params, double[] hypers) {
if (model(hypers) != null) return null;
ModelBuilder mb = getBuilder(params, hypers);
mb.trainModel();
return mb;
}