numerous.sdk.connect.writer module#

Functionality related to writing data series and the Writer.

exception numerous.sdk.connect.writer.TagsNotAllowed(tags)#

Bases: Exception

An error raised when written tags are invalid according to the data series.

class numerous.sdk.connect.writer.Writer(spm_stub, job_identity, start_time, execution_id, max_size_bytes, flush_margin_bytes)#

Bases: object

The Writer enables writing rows and series to the server. It manages a buffer, which is automatically flushed when it is full.

close(hibernating)#

Close the Writer, flushing the buffer.

flush()#

Flushes the buffer of the writer.

Return type:

None

row(time, data, flush=False)#

Write a row of data into the data series.

Each row written to the Writer must have same keys as previously written series or rows.

Parameters:

  • time (Union[float, datetime]) – The time, at which the data is indexed. If a float value, it is interpreted as an offset in seconds since the start time of the job, otherwise if it is a datetime, it is interpreted as the time within the job runtime.

  • data (dict[str, float]) – The row of data to write into the data stream.

  • flush (bool) – If true, will flush the buffer of the writer upon writing.

Raises:

TagsNotAllowed – If tags that have not been previously written, are written.

Return type:

None

Writing rows can be used alongside numerous.sdk.JobClient.read_inputs(), which will also iterate over the jobs runtime:

>>> for t, data in job_client.read_inputs(step=3600.0):
...     job_client.writer.row(t, {"output": data["input1"] * 2})

The type of the time values does not have to be the same type for each call:

>>> job_client.job_time.start
datetime(2023, 6, 1, 12, 0 ,0)
>>> job_client.writer.row(datetime(2023, 6, 1, 12, 0, 0), {"value": 1.0})
>>> job_client.writer.row(3600.0, {"value": 2.0})
>>> job_client.writer.row(datetime(2023, 6, 1, 14, 0, 0), {"value": 4.0})
series(times, data, flush=False)#

Write a series of data to the data stream.

Each series written to the Writer must have same keys as previously written series or rows.

Parameters:

  • times (Sequence[Union[float, datetime]]) – The time values of the series, like for row(), they can be either offsets from the start time, or a point in time.

  • data (dict[str, list[float]]) – The data series to write. The keys are validated against previously written keys.

  • flush (bool) – If true, will flush the buffer of the writer upon writing.

Raises:

TagsNotAllowed – If tags that have not been previously written, are written.

Return type:

None