Reputation: 6615
I think fairly simple question - I'm looking to understand how I can extract various bits of metadata out of a h2o model I built in python?
In R I can run the following to get this sort of information.
my_h2o_model@algorithm
my_h2o_model@parameters
my_performance = h2o.performance(my_h2o_model)
my_performance@metrics$thresholds_and_metric_scores # awesome see all classifier metrics at each threshold
In python if I try
my_h2o_model['algorithm']
or
my_h2o_model[0]
perf=model_performance(my_h2o_model)
perf['metrics']['thresholds_and_metric_scores] # !!!!
TypeError: 'H2OGradientBoostingEstimator' object is not subscriptable
How can get this information out of the models in python??
Upvotes: 0
Views: 242
Reputation: 8819
You have a typo -- you're missing a single quote after thresholds_and_metric_scores
.
perf['metrics']['thresholds_and_metric_scores']
Upvotes: 1