Skip to content

on_boarding.md

Latest commit

 

History

History
118 lines (77 loc) · 9.02 KB

on_boarding.md

File metadata and controls

118 lines (77 loc) · 9.02 KB
graph LR
    Data_Ingestion_Conversion["Data Ingestion & Conversion"]
    Core_Inference_Data_Structure["Core Inference Data Structure"]
    Statistical_Analysis_Diagnostics["Statistical Analysis & Diagnostics"]
    High_Level_Plotting_API["High-Level Plotting API"]
    Matplotlib_Plotting_Backend["Matplotlib Plotting Backend"]
    Bokeh_Plotting_Backend["Bokeh Plotting Backend"]
    Core_Utilities_Helpers["Core Utilities & Helpers"]
    Data_Ingestion_Conversion -- "produces" --> Core_Inference_Data_Structure
    Core_Inference_Data_Structure -- "provides data to" --> Statistical_Analysis_Diagnostics
    Core_Inference_Data_Structure -- "provides data to" --> High_Level_Plotting_API
    High_Level_Plotting_API -- "delegates rendering to" --> Matplotlib_Plotting_Backend
    High_Level_Plotting_API -- "delegates rendering to" --> Bokeh_Plotting_Backend
    Core_Utilities_Helpers -- "supports" --> Data_Ingestion_Conversion
    Core_Utilities_Helpers -- "supports" --> Core_Inference_Data_Structure
    Core_Utilities_Helpers -- "supports" --> Statistical_Analysis_Diagnostics
    Core_Utilities_Helpers -- "supports" --> High_Level_Plotting_API
    Core_Utilities_Helpers -- "supports" --> Matplotlib_Plotting_Backend
    Core_Utilities_Helpers -- "supports" --> Bokeh_Plotting_Backend
    click Data_Ingestion_Conversion href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/arviz/Data_Ingestion_Conversion.md" "Details"
    click Core_Inference_Data_Structure href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/arviz/Core_Inference_Data_Structure.md" "Details"
    click Statistical_Analysis_Diagnostics href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/arviz/Statistical_Analysis_Diagnostics.md" "Details"
    click High_Level_Plotting_API href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/arviz/High_Level_Plotting_API.md" "Details"
    click Matplotlib_Plotting_Backend href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/arviz/Matplotlib_Plotting_Backend.md" "Details"
    click Bokeh_Plotting_Backend href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/arviz/Bokeh_Plotting_Backend.md" "Details"
Loading

CodeBoardingDemoContact

Details

ArviZ is structured around a central InferenceData object, which serves as the universal container for Bayesian inference results. Data from various probabilistic programming frameworks is first ingested and converted into this standardized InferenceData format. Once in this core structure, the data can be subjected to comprehensive statistical analysis and diagnostics, or visualized through a high-level plotting API. This plotting API abstracts away the underlying rendering mechanisms, delegating to specific backends like Matplotlib or Bokeh for static or interactive visualizations, respectively. A collection of core utilities and helper functions provides foundational support across all these components, ensuring consistent data handling, selection, and configuration.

Data Ingestion & Conversion [Expand]

Handles the intake and standardization of raw output from various probabilistic programming frameworks into ArviZ's InferenceData format.

Related Classes/Methods:

Core Inference Data Structure [Expand]

The central, immutable data container (InferenceData object) that stores and manages all aspects of Bayesian inference results, built upon xarray.Dataset.

Related Classes/Methods:

Statistical Analysis & Diagnostics [Expand]

Provides a comprehensive suite of functions for assessing MCMC convergence, mixing, and performing various statistical analyses like model comparison (LOO, WAIC), posterior summaries, and HDI calculations.

Related Classes/Methods:

High-Level Plotting API [Expand]

Offers a unified, user-facing API for generating a wide range of plots from InferenceData objects, abstracting away backend-specific implementations.

Related Classes/Methods:

Matplotlib Plotting Backend [Expand]

Contains the concrete implementations for rendering ArviZ plots using the Matplotlib library, producing static visualizations.

Related Classes/Methods:

Bokeh Plotting Backend [Expand]

Contains the concrete implementations for rendering ArviZ plots using the Bokeh library, enabling interactive web-based visualizations.

Related Classes/Methods:

Core Utilities & Helpers

A collection of general-purpose helper functions, data selection utilities, labeling mechanisms, and configuration management used across various ArviZ components.

Related Classes/Methods:

FAQ