partitionnode properties
Last updated: Feb 11, 2025
The Partition node generates a partition field, which splits
the data into separate subsets for the training, testing, and validation stages of model
building.
Example
node = stream.create("partition", "My node") node.setPropertyValue("create_validation", True) node.setPropertyValue("training_size", 33) node.setPropertyValue("testing_size", 33) node.setPropertyValue("validation_size", 33) node.setPropertyValue("set_random_seed", True) node.setPropertyValue("random_seed", 123) node.setPropertyValue("value_mode", "System")
properties |
Data type | Property description |
---|---|---|
|
string | Name of the partition field generated by the node. |
|
flag | Specifies whether a validation partition should be created. |
|
integer | Percentage of records (0–100) to be allocated to the training partition. |
|
integer | Percentage of records (0–100) to be allocated to the testing partition. |
|
integer | Percentage of records (0–100) to be allocated to the validation partition. Ignored if a validation partition is not created. |
|
string | Label for the training partition. |
|
string | Label for the testing partition. |
|
string | Label for the validation partition. Ignored if a validation partition is not created. |
|
|
Specifies the values used to represent each partition in the data. For example, the training
sample can be represented by the system integer , the label
, or a combination of the two, . |
|
Boolean | Specifies whether a user-specified random seed should be used. |
|
integer | A user-specified random seed value. For this value to be used,
must be set to . |
|
Boolean | Specifies whether to use SQL pushback to assign records to partitions. |
|
Specifies the input field used to ensure that records are assigned to partitions in a random
but repeatable way. For this value to be used, must be set to
. |
Was the topic helpful?
0/1000