Skip to content

Reporting objects

This tutorial will show you how pipestat can report not just primitive types, but structured results as objects.

First create a pipestat.PipestatManager object with our example schema:

import pipestat
psm = pipestat.PipestatManager(
Initialize PipestatBackend
Initialize FileBackend

Here, we're pointing to a remote schema_path. Let's take a look at the schema object. You can see a preview of it here at the schema server: We can also see what it looks like in Python code:

ParsedSchemaDB (refget)
 Project-level properties:
 - None
 Sample-level properties:
 -  value : {'type': 'string', 'description': 'Value of the object referred to by the key'}
 -  mydict : {'type': 'object', 'description': 'Can pipestat handle nested objects?'}
 Status properties:
 - None

This schema defines two sample-level variables: value is a string, and mydict is an object. Let's see how to report a mydict:{"mydict": {"toplevel": {"value": "456"}}}, force_overwrite=True)
["Reported records for 'sample1' in 'refget' :\n - mydict: {'toplevel': {'value': '456'}}",
 "Reported records for 'sample1' in 'refget' :\n - pipestat_created_time: 2024-02-21 20:52:20",
 "Reported records for 'sample1' in 'refget' :\n - pipestat_modified_time: 2024-02-21 20:52:20"]

And now we can retrieve those results:

psm.retrieve_one("sample1", "mydict")
{'toplevel': {'value': '456'}}

Is there a way to retrieve the values within? Something like psm.retrieve_one("sample1", "mydict.toplevel.value") ? I can't find any docs about that.