@Singleton @Path(value="/classificationDistribution") public final class ClassificationDistribution extends AbstractOryxResource
Responds to a GET request to /classificationDistribution/[datum].
Like Predict but this
returns not just the most probable category,
but all categories and their associated probability. It is not defined for
regression problems and returns an error.
Default output is CSV format, containing category,probability per line.
JSON format can also be selected by an appropriate Accept header. It returns
an array of probabilities, each of which has an "id" and "value" entry, like
[{"id":"I2","value":0.141348009071816},...]
INPUT_PRODUCER_KEY, MODEL_MANAGER_KEY| Constructor and Description |
|---|
ClassificationDistribution() |
| Modifier and Type | Method and Description |
|---|---|
List<IDValue> |
get(String datum) |
check, check, checkExists, checkNotReadOnly, getServingModel, isReadOnly, maybeBuffer, maybeBuffer, maybeDecompress, parseMultipart, sendInputgetInputProducer, getServingModelManager, init@GET
@Path(value="{datum}")
@Produces(value={"text/plain","text/csv","application/json"})
public List<IDValue> get(@PathParam(value="datum")
String datum)
throws OryxServingException
OryxServingExceptionCopyright © 2014–2018. All rights reserved.