Skip to main content

Hello


Is it possible to specify the “Request id” to facilitate experiment tracking?



Thanks

Hi,


the request ID is generated automatically. However, you can customize the path where the results will be stored using defaultTilePath. For example, to store everything inside the prefix “my-experiment-id/<request-id>”, include this when you create the batch request:


{
...,
"output": {
"defaultTilePath": "s3://my-bucket/my-experiment-id"
}
}

If you do not want to have the request ID appended, you can provide a templated path that tells the service how to organize the output. For example, to have everything delivered directly into the “my-experiment-id” prefix with no subfolders:
"defaultTilePath": "s3://my-bucket/my-experiment-id/<tileName>.tif"

With templates there are some caveats about already existing files and uniqueness, so please check the API reference for details.


Thanks!


And what about the case of batch statistical API?

There is no “defaultTilePath”


The API reference for statistical batch was indeed lacking this info, we have just updated it. Please note that the Statistical Batch service is still in Beta, thus the API might change in the future.


To put all output JSONs into a single folder named “my-experiment-id” using the “identifier” property of the features as filenames you would use:


{
...,
"output": {
"s3": {
"url": "s3://my-bucket/my-experiment-id/<IDENTIFIER>.json",
...
}
}
}

Reply