LinkML-Store documentation

An AI-ready data management and integration platform. LinkML-Store provides an abstraction layer over multiple different backends (including DuckDB, MongoDB, and local filesystems), allowing for common query, index, and storage operations.

Contents:

• About
  • Installation
  • Quickstart
  • Data Model
  • Adapters
  • Indexing
  • Metadata and Configuration
• Tutorials
  • Tutorial: Using the Command Line Interface
    • Top level command
    • Inserting objects from a file
    • Querying
    • Describing the data set
    • Facet Counts
    • Search
    • Indexing using an LLM (OPTIONAL)
    • Schemas
    • Configuration Files and Explicit Schemas
    • Validation
    • Inference
    • Inference using statistical models
  • Tutorial: Using the Python API
    • Create a Client object
    • Attach and populate a DuckDB in-memory database
    • Querying
    • Faceting
    • Plotting
    • Complex objects
    • Schemas
    • Asserting schemas
    • Other adapters
    • Indexing
• How-To Guides
  • How to use MongoDB with LinkML-Store
    • Creating a client and attaching to a database
    • Creating a collection
    • Preparing data to load
    • Inserting objects
    • Queries
    • Facet counts
    • Creating an LLM embedding index
  • How to use Semantic Search
    • Background
    • Obtaining upstream files
    • Storing the JSON
    • Creating an LLMIndexer
    • Searching using the index
    • How it works
  • How to query the Monarch-KG
  • How to query Solr using the Command Line
    • Specifying a Solr backend on the command line:
    • Using a configuration
    • Search
  • How to use Neo4J with LinkML-Store
    • Running this how-to guide interactively.
    • Using a schema
    • Create Node and Edge collections
    • Adding Data
    • Visualization
    • Validation
    • Loading Data from a file
    • Queries
    • Facet Counts
    • Indexing via LLM embeddings
  • How to Check Referential Integrity
    • Inserting invalid data
  • Example: Storing an ontology
  • How to index Phenopackets with LinkML-Store
    • Use pystow to download phenopackets
    • Creating a client and attaching to a database
    • Creating a collection
    • Inserting objects into the store
    • Check contents
    • Query faceting
    • Semantic Search
    • Semantic Search
    • Validation
    • Command Line Usage
    • Inference
  • How to index GO-CAMs with LinkML-Store
    • Creating a client and attaching to a database
    • Creating a collection
    • Inserting objects into the store
    • Check contents
    • Query faceting
    • Semantic Search
    • Semantic Search
    • Validation
    • Command Line Usage
  • How to predict missing data
    • The Infer Command
    • Training and Inference
    • Saving the Model
    • Exporting models to explainable visualizations
    • Generating a rule-based model
    • More advanced ML models
  • Perform RAG Inference
    • Loading the data into duckdb
    • Inferring a specific field
    • Inferring a whole object
    • Inferring from multiple fields
    • RAG configuration - using a different model
    • Persisting the RAG model
    • Evaluation
    • How RAG indexing works under the hood
    • Configuring the training/test split
    • Extraction tasks
• Reference
  • API
    • Client
    • Database
    • Collection
  • Adapters
    • Subpackages
• Indices and tables
• Frequently Asked Questions
  • General
    • What is this project?
    • Is this a database engine?
    • Do I need to know LinkML to use this?
    • Can I use the command line?
    • Can I use the Python API?
    • Can I use a web API?
    • Can I use a web UI?
    • What is CRUDSI?
    • Is this an AI/Machine Learning/LLM/Vector database platform?
    • Is linkml-store production ready?
    • Are there tutorials?
  • Installation
    • How do I install linkml-store?
    • As a developer, how do I install linkml-store?
  • Backend Integrations
    • What is a database integration?
    • Does linkml-store support DuckDB?
    • Does linkml-store support MongoDB?
    • Does linkml-store support Neo4J?
    • Does linkml-store support Solr?
    • Can I use linkml-store with my favorite triplestore?
  • Data model
    • What is the data model in linkml-store?
  • Search
    • Can I use LLM vector embeddings for search?
    • Do I need to use an LLM for search
  • Validation
    • Does linkml-store provide validation?
    • Does linkml-store provide referential integrity validation?
  • Inference
    • What is inference in linkml-store?
    • How do I do standard ML inference?
    • How do I do LLM inference?
    • How do I do rule-based inference?
  • Contributing
    • How do I contribute to linkml-store?

Indices and tables

• Index

• Module Index

• Search Page