Skip to content

ISARIC Reusable Analytical Pipelines for Infectious Diseases (RAPID)

RAPID Methodology: Reusable Analytical Pipelines for Infectious Diseases

Welcome to the official documentation for the RAPID Methodology (Reusable Analytical Pipelines for Infectious Diseases).

RAPID is a robust analytical structure, adapted by ISARIC, which integrates software engineering principles to formalize and standardize the entire analytical workflow in clinical research concerning infectious diseases.

Our core objective is to ensure that analyses are Reproducible, Transparent, Efficient, and Robust.


The RAPID Methodology Flow

The RAPID methodology transforms a clinical research question and a curated dataset into a transparent and validated analysis with clinical relevance. Here's the updated Mermaid block — just replace the existing one:

graph LR
    subgraph INPUT
        A[ Research<br/>Question]
        B[ Curated<br/>Dataset]
    end

    subgraph PROCESS[" RAPID Methodology"]
        C1[1  Cleaning] --> C2[2 Preprocessing]
        C2 --> C3[3 Modeling]
        C3 --> C4[4 Evaluation]
        C4 --> C5[5 Validation]
        C5 --> C6[6 Visualization]
        C6 -.-> C2
    end

    subgraph OUTPUT
        D[ Validated<br/>Analysis]
    end

    A --> C1
    B --> C1
    C6 --> D

The 6-Step RAPID Workflow

Step Phase Mandatory Description
1 Cleaning Remove outliers, handle duplicates, standardize formats
2 Preprocessing MICE imputation, encoding, normalization
3 Modeling Apply statistical models (GLM, Survival, MICE)
4 Evaluation Performance metrics, concordance index, Brier score
5 Validation Bootstrapping, cross-validation
6 Visualization Survival curves, forest plots, hazard ratios

Iterative Process

Visualization can occur at any stage. The dashed arrows indicate that researchers may return to previous steps to refine the model based on evaluation results.

Full Methodology Document

For complete details on the RAPID methodology, including theoretical foundations and design decisions:

Download the RAPID Methodology (PDF)


Start Your Analysis

Installation

Install the RAPID package directly from GitHub:

pip install git+https://github.com/noispuc/isaric3.0-rapid.git

For detailed instructions and alternative methods, see the Installation Guide

Choose Your Path

You want to... Go to...
Understand why RAPID exists Why RAPID?
Install the package Installation Guide
Get started quickly Quickstart Guide
Learn statistical theory + practice User Guide
Quick reference (parameters & methods) Quick Guide
See real-world examples Examples Gallery

Primary Audience

The RAPID methodology is designed to support two main groups of users:

  • End-users: Researchers, analysts, and domain experts who apply RAPID tools to generate insights and evidence using established workflows.
  • Contributors and Developers: Data scientists, software engineers, and technical researchers who seek to extend RAPID's capabilities, including developing new functionalities or refining existing components.

See our Contributing Guidelines to get involved.


Contact & Support