Episodes_BatchSampleEpisodeSeries
const url = 'https://junction-api.qualiastudios.dev/v1/accounts/example/datasets/example/episodes:batchSampleSeries';const options = {method: 'GET', headers: {'X-API-Key': '<X-API-Key>'}};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request GET \ --url https://junction-api.qualiastudios.dev/v1/accounts/example/datasets/example/episodes:batchSampleSeries \ --header 'X-API-Key: <X-API-Key>'Several episodes’ series at once, each as SampleEpisodeSeries samples it: what a dataset’s trajectories read. The takes are read together, so the artifacts they share are opened once and their reads overlap; each take still answers on its own, so one that cannot be read fails alone. Up to 50 takes, each an episode of the parent, and 20000 samples (the takes times max_points).
Authorizations
Section titled “Authorizations ”Parameters
Section titled “ Parameters ”Path Parameters
Section titled “Path Parameters ”The account id.
The dataset id.
Query Parameters
Section titled “Query Parameters ”The takes, 1 to 50, each an episode of the parent: accounts/{account}/datasets/{dataset}/episodes/{episode}. Answered in this order.
As SampleEpisodeSeriesRequest.max_points, for every take. The takes times max_points (400 when unset) are at most 20000: 50 takes at 400, or 10 at 2000.
As SampleEpisodeSeriesRequest.streams, for every take.
As SampleEpisodeSeriesRequest.version, for every take.
Responses
Section titled “ Responses ”OK
Response for Episodes.BatchSampleEpisodeSeries.
object
One per take asked, in the order asked.
One take’s answer.
object
Accounts/{account}/datasets/{dataset}/episodes/{episode}
Its series, as SampleEpisodeSeries answers; unset where it failed.
object
Seconds from the first frame sampled, one per sample: a sample’s second of the episode’s recording is this plus window_start_seconds.
Every channel the recording carries, on time_seconds.
One channel: one component of the state or the action vector.
object
Which vector.
Its positional key, “state[3]” or “action[3]”; the dataset’s robot interface names the positions. A TCP channel is named, “left.tcp.x”.
One value per sample; 0 where valid is false.
Per sample: false where the recording holds no finite value.
The recording’s frames per second.
Frames sampled: the take’s, or what its applied trim keeps.
True when the take is in no published version yet (a split child before its publish): it has no channels to sample.
Where the first frame sampled sits in the episode’s recording, in seconds of the recording (the clock its trim, subtasks and markers are in), on that frame: 0 until a trim is applied, the trim’s start once one is; a split’s piece is a recording of its own, so 0. Unset with unpublished.
Where the frames sampled end in the recording, in the same seconds: the end of the last one (window_start_seconds plus frame_count / fps). Unset with unpublished.
Where it failed, the gRPC status code SampleEpisodeSeries would have answered for this take, as google.rpc.Code numbers it (5 NOT_FOUND, 9 FAILED_PRECONDITION, …); 0 where it was read.
Where it failed, why; empty otherwise.
Example
{ "results": [ { "series": { "channels": [ { "stream": "STREAM_UNSPECIFIED" } ] } } ]}default
Section titled “default ”Default error response
A refusal from the API itself.
object
The gRPC code name in upper case, e.g. NOT_FOUND.
Whether the same call may succeed if retried.
Empty, except when a dataset copy or move is refused: then one google.rpc.ErrorInfo with the refusal’s reason code, its domain and, in metadata, the datasets it names.
object
A refusal from the gateway before the call reaches the API, for example 401 bad or missing API key, 403 key lacks the scope or environment closed, 413 body too large.
object
What went wrong.
Example generated
{ "code": "example", "message": "example", "retryable": true, "details": [ {} ]}