Gets memory metrics and prediction metrics for a deployed PMML model from Zementis Server.
get_model_metrics(model_name, ...)
model_name | Name of the PMML model whose metrics are fetched from the server. |
---|---|
... | Additional arguments passed on to the underlying HTTP method.
This might be necessary if you need to set some curl options explicitly
via |
A list with the following components:
model_name
A length one character vector containing the model_name
prediction_metrics
A data frame containing prediction-related
metrics for model_name
. The information contained in prediction_metrics
differs between regression and classification models.
memory_metrics
A data frame containing memory-related metrics
for model_name
expressed in MB.
If no predictions have been calculated for model_name
thus far on Zementis Server,
prediction_metrics
won't be included in the response list.
If the model is deactivated while get_model_metrics()
is called, the return list
neither includes memory_metrics
nor prediction_metrics
.
The HTTP endpoint accessed by get_model_metrics()
is only available for Zementis Server 10.3 or higher.
See vignette("model-metrics") for more details on that function and for best practices how to visualize the
different model metrics of your predictive models.
if (FALSE) { # Some prep work iris_lm <- lm(Sepal.Length ~ ., data = iris) iris_pmml <- pmml::pmml(iris_lm, model.name = "iris_model") upload_model(iris_pmml) # only includes memory metrics get_model_metrics("iris_model") predict_pmml_batch(iris[23:33, ], "iris_model") # includes memory and prediction metrics get_model_metrics("iris_model") }