RandomSeed
A number to use as a seed for the cluster sampling algorithm.
IDOL Category Component uses a sample of documents to build the seeds around which clusters are formed. It usually selects a sample randomly, which means that the criteria vary over subsequent runs, and two snapshots of the same data might be slightly different. Set RandomSeed
to ensure that the samples are the same, so that you can exactly reproduce a snapshot of the same data.
Set RandomSeed
to 0
or above to use a sampling algorithm that does not vary over subsequent runs.
NOTE: Different values produce snapshots that are slightly different from each other, but all values produce snapshots that can be reproduced.
Set RandomSeed
to -1
to use the default setting (different snapshots of the same data might vary).
NOTE: The setting that you make here also applies to the algorithm that is used to produce 2-D maps. That is, a value of 0
or above ensures that 2-D maps produced from the same data are always identical.
Type: | Long |
Default: | -1 |
Required: | No |
Configuration Section: | Server |
Example: | RandomSeed=10
|
See Also: |