Skip to main content

A suite of tools for interacting with Zegami through Python.

Project description

Zegami Python SDK

An SDK and general wrapper for the lower level Zegami API for Python. This package provides higher level collection interaction and data retrieval.

Getting started

Grab this repo, open the script, and load an instance of ZegamiClient into a variable.

from zegami_sdk.client import ZegamiClient

zc = ZegamiClient(username=USERNAME, password=PASSWORD)


The client operates using a user token. By default, logging in once with a valid username/password will save the acquired token to your home directory as zegami.token. The next time you need to use ZegamiClient, you may call zc = ZegamiClient() with no arguments, and it will look for this stored token.

Example Usage

zc = ZegamiClient()


To see your available workspaces, use:


You can then ask for a workspace by name, by ID, or just from a list

all_workspaces = zc.workspaces
first_workspace = all_workspaces[0]



# Note the ID of a workspace
my_workspace = zc.get_workspace_by_id(id)



# Note the name of a collection
coll = my_workspace.get_collection_by_name(name_of_collection)

You can get the metadata in a collection as a Pandas DataFrame using:

rows = coll.rows

This data can then be modified or augmentated and added back to the collection using:


You can get the images of a collection using:

first_10_img_urls = coll.get_image_urls(list(range(10)))
imgs = coll.download_image_batch(first_10_img_urls)


If a collection contains multiple image sources, these can be seen using:


Many operations require specifying which image source should be used. This can be specified by index or name for most functions.

first_10_source2_img_urls = coll.get_image_urls(list(range(10)), source=2)

# To see the first of these:

Using with onprem zegami

To use the client with an onprem installation of zegami you have to set the home keyword argument when instantiating ZegamiClient.

zegami_config = {
  'username': <user>,
  'password': <password>,
  'home': <url of onprem zegami>,
  'allow_save_token': True,

zc = ZegamiClient(**zegami_config)

If your onprem installation has self-signed certificates you can disable SSL verification using the environment variable ALLOW_INSECURE_SSL before running the python.




WARNING! You should not need to set this when using the SDK for cloud zegami

In Development

This SDK is in active development. Features are actively being developed according to user feedback. Please share your suggestions or fork this repository and feel free to raise a PR

Developer Conventions

Keeping the SDK easy and fluent to use externally and internally is crucial. If contributing PRs, some things to consider:


MOST IMPORTANT - Zegami has concepts used internally in its data engine, like 'imageset', 'dataset'. Strive to never require the user to have to know anything about these, or even see them. If the user needs an image, they should ask for an image from a concept they ARE expected to understand like a 'collection' or a 'workspace'. Anything obscure should be hidden, for example: _get_imageset(), so that auto-suggestions of a class will always contain relevant and useful methods/attribs/properties.


Avoid ambiguous parameters. Use the best worded, lowest level parameters types for functions/methods. Give them obvious names. Any ambiguity or unobvious parameters MUST be described in detail in the docstring. Avoid parameters like 'target' or 'action', or describe them explicitly. If an instance is needed, describe how/where that instance should come from.


If you expect an RGB image, check that your input is an array, that its len(shape) == 3, that shape[2] == 3, and throw an exception to clearly feed back to the user what has wrong. The message should help the user to solve the problem for themselves.


Do not ask for more information than is already obtainable. A source knows its parent collection, which knows how to get its own IDs and knows the client. A method never needs to reference a source, the owning collection, and the client all together. Moreover, these chains should have sensible assertions and checks built in, and potentially property/method-based shortcuts (with assertions).


Use sensible defaults wherever possible for minimal effort when using the SDK. V1 collections typically use source=None, while V2 collections use source=0. This allows a user with an old/new (single source) collection to never even have to know what a source is when fetching images.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

zegami-sdk-0.4.9.tar.gz (36.9 kB view hashes)

Uploaded Source

Built Distribution

zegami_sdk-0.4.9-py3-none-any.whl (42.1 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page