CorePlexMLClient
The CorePlexMLClient is the entry point for all SDK
operations. It initializes the HTTP transport and exposes every API resource
as a typed attribute.
Constructor
CorePlexMLClient(base_url="http://localhost:8888", api_key=None, timeout=30)
Parameters
Parameter |
Type |
Default |
Description |
|---|---|---|---|
|
|
|
CorePlexML server URL. Include the scheme ( |
|
|
|
API key for Bearer authentication. Generate one from the web UI. |
|
|
|
HTTP request timeout in seconds. |
Resource Attributes
Every resource is initialized once during client construction and is accessible as a read-only attribute.
Attribute |
Type |
Description |
|---|---|---|
|
Create, list, update, delete projects and manage members. |
|
|
Upload, list, download datasets and inspect columns. |
|
|
Create AutoML experiments and poll for completion. |
|
|
List trained models, get details, and make predictions. |
|
|
Deploy models, make predictions, promote/rollback. |
|
|
Generate and download PDF reports. |
|
|
PII detection and data transformation. |
|
|
Train generative models and produce synthetic data. |
|
|
What-If Analysis – scenario-based model exploration. |
Connection Configuration
The SDK uses the requests library internally. The base URL is normalized
(trailing slashes are stripped), and the API key is sent as a
Authorization: Bearer <key> header on every request.
# HTTPS with a long timeout for large uploads
client = CorePlexMLClient(
base_url="https://ml.example.com",
api_key="cp_ab12cd34.your-secret-key",
timeout=120,
)
# Local development (no auth)
local_client = CorePlexMLClient()
Authentication
API keys are created in the CorePlexML web UI under Settings > API Keys.
Each key has a prefix and a secret separated by a dot (e.g.,
cp_ab12cd34.your-secret-key). Keys are scoped to a user account and inherit that
user’s project permissions.
If no api_key is provided, requests are sent without an Authorization
header. This may work for unauthenticated local development endpoints but will
fail against production servers.
from coreplexml import CorePlexMLClient, AuthenticationError
client = CorePlexMLClient(
base_url="https://ml.example.com",
api_key="cp_ab12cd34.your-secret-key",
)
try:
client.projects.list()
except AuthenticationError:
print("API key is invalid or expired. Generate a new one in Settings.")
Autodoc Reference
- class coreplexml.CorePlexMLClient(base_url='http://localhost:8888', api_key=None, timeout=30)[source]
Bases:
objectTop-level client for the CorePlexML Python SDK.
Provides access to all API resources through typed sub-clients.
- Parameters:
Example:
from coreplexml import CorePlexMLClient client = CorePlexMLClient("https://your-domain.com", api_key="your-key") projects = client.projects.list()
- projects
Manage projects.
- datasets
Upload, list, and manage datasets.
- experiments
Create and monitor AutoML experiments.
- models
Get model info and make predictions.
- deployments
Deploy models and manage production endpoints.
- reports
Generate and download reports.
- privacy
Privacy Suite – PII detection and data transformation.
- synthgen
Synthetic data generation with CTGAN/CopulaGAN/TVAE/Gaussian Copula.
- studio
What-If Analysis – scenario comparison.