Python SDK

Installation

Install the package from Pypi using any package manager. Examples:

• poetry add marpledata

• uv add marpledata

• pip install marpledata

Usage

Read the full documentation on Pypi: https://pypi.org/project/marpledata/

Example: pushing a file

import time
from marple import DB

# create a datastream and API token in the Marple DB web application
DATASTREAM = 'Car data'
API_TOKEN = '<your api token>'
# optional, if you are not on db.marpledata.com
API_URL = '<e.g. db.customer.marpledata.com/api/v1 - link to db, ending in /api/v1>'

db = DB(API_TOKEN, API_URL)

dataset_id = db.push_file(DATASTREAM, "tests/example_race.csv", metadata={"driver": "Mbaerto"})
print(db.get_status(DATASTREAM, dataset_id))

Migration from Marple Files

Using the newest features for the SDK requires version 2.0.0 or higher. Check your resolved packages (requirements.txt, poetry.lock, ...) and upgrade if necessary.

There are two major changes to keep in mind while migrating:

1. There are two products now instead of one. But 90% of data-related functionality is in Marple DB.

2. When uploading a new file, you no longer need to provide the plugin and import options. Instead, you need to configure a datastream which takes care of this. Read more here:

Configuration

The table below shows functionality from Marple Files and their counterparts in the new SDK:

Marple Files (Deprecated)	Marple DB	Marple Insight
/library/upload or m.upload_data_file(...)	db.push_file(...) ℹ️ plugin is now configured by datastream (once)	-
m.check_import_status(...)	db.get_status(...)	-
/sources/info	db.get_datasets(...)	-
/library/folder/contents	🚧 Upcoming	-
/export	db.download_original(...)	insight.export_mdb(...) Supported: mat, h5 🚧 Upcoming: csv
m.get_link(...)	-	Use dynamic share links