pd4castr Documentation

Create, test, and publish forecasting models — or integrate via API.

Get started with Model Authoring API SDK Platform Guide

Validate and publish your models

Initialize new model projects, validate them locally, then publish them to the pd4castr platform from the pd4castr CLI.

$ npm install -g @pd4castr/cli
$ pd4castr login
$ pd4castr init
$ pd4castr publish

Integrate with the pd4castr API

Manage your models and query your model runs via our API with the pd4castr Python SDK.

from pd4castr_api_sdk import Client

client = Client(client_id="...", client_secret="...")

models = client.get_models()
model = client.get_model(model_id=models[0].id)

Model Authoring

Learn the fundamentals and publish your first model.

Project structure
Configuration
Inputs/Outputs
Local vs platform runs
Author your first modelModel configuration

API SDK

Integrate with the pd4castr API using our Python SDK.

SDK setup
Common operations
Authentication
Error Handling
Quick start with the SDKSDK Reference

Platform Guide

Learn the features of pd4castr and how the platform works in detail.

Platform overview
Forecasts
Model runs
Sensitivities
Read the overviewExplore data

Quick Start

Get up and running with pd4castr in 5 steps

1

Install CLI

Get the pd4castr command-line tools

Installation guide
2

Authenticate

Set up your credentials and connect

Auth setup
3

Fetch inputs

Pull forecast data to work with

Data fetching
4

Test locally

Run and validate your model

Local testing
5

Publish

Deploy your model to production

Publishing