Logo

iMML

  • Overview
    • Key features
    • Installation
    • Usage
    • Free software
    • Contribute
    • Help us grow
  • Installation
    • Instructions
      • Using pip
      • Using uv
      • Manually
      • Dependencies
    • Optional dependencies
    • OS Requirements
    • Testing
    • Documenting
  • Algorithm selection guide
    • How to install an additional module
    • How to install extra dependencies
      • Extra dependencies when using "octave" module
      • Extra dependencies when using "r" module
  • Tutorials
    • Statistics and interaction structure of a multi-modal dataset
      • Step 1: Import required libraries
      • Step 2: Create or load a multi-modal dataset
      • Step 3: Compute PID statistics (Redundancy, Uniqueness, Synergy)
      • Step 4: Visualize the PID as a Venn-like diagram
        • Interpreting PID results
        • Working with more than two modalities
      • Step 5: Summarize the dataset
      • Conclusion
    • Modality-wise missing data simulation (Amputation)
      • Step 1: Import required libraries
      • Step 2: Load the dataset
      • Step 3: Simulate missing data
      • Step 4: Compare amputation mechanisms
      • Step 5: Vary the missingness rate
      • Conclusion
    • Clustering a multi-modal dataset
      • Step 1: Import required libraries
      • Step 2: Load the dataset
      • Step 3: Clustering
      • Step 4: Simulate missing data (Amputation)
      • Step 5: Clustering with missing data
      • Step 6: Benchmarking
      • Summary of results
      • Conclusion
    • Retrieval on a vision–language dataset (flickr30k)
      • Step 0: Prerequisites
      • Step 1: Import required libraries
      • Step 2: Prepare the dataset
      • Step 3: Simulate missing modalities
      • Step 4: Generate the memory bank
      • Step 5: Retrieve
      • Step 6: Visualize the retrieved instances
      • Summary of results
      • Conclusion
    • Impute modality- and feature-wise incomplete multi-modal data
      • Step 1: Import required libraries
      • Step 2: Load the dataset
      • Step 3: Impute missing data
      • Step 4: Benchmark imputation accuracy
      • Summary of results
      • Conclusion
    • Classify an incomplete vision–language dataset (Oxford‑IIIT Pets) with deep learning
      • Step 0: Prerequisites
      • Step 1: Import required libraries
      • Step 2: Prepare the dataset
      • Step 3: Simulate missing modalities
      • Step 4: Generate the prompts using a retriever
      • Step 5: Training the model
      • Step 6: Advanced Usage: Track Metrics During Training
      • Step 7: Evaluation
      • Summary of results
      • Conclusion
    • Dimensionality reduction: Feature extraction and feature selection
      • Step 0: Prerequisites
      • Step 1: Import required libraries
      • Step 2: Define plotting functions
      • Step 3: Load the dataset
      • Step 4: Apply feature selection and feature extraction
      • Step 6: Analyzing an incomplete multi-modal dataset
      • Summary of results
      • Conclusion
  • API Reference
    • Ampute
      • Amputer
      • Remove Modalities
    • Classify
      • Missing Modalities in Multimodal healthcare data (M3Care)
        • M3Care
          • M3Care.training_step()
          • M3Care.validation_step()
          • M3Care.test_step()
          • M3Care.predict_step()
          • M3Care.configure_optimizers()
      • MUtual-conSistEnt graph contrastive learning (MUSE)
        • MUSE
          • MUSE.training_step()
          • MUSE.validation_step()
          • MUSE.test_step()
          • MUSE.predict_step()
          • MUSE.configure_optimizers()
      • Retrieval-AuGmented dynamic Prompt Tuning (RAGPT)
        • RAGPT
          • RAGPT.training_step()
          • RAGPT.validation_step()
          • RAGPT.test_step()
          • RAGPT.predict_step()
          • RAGPT.configure_optimizers()
    • Cluster
      • Doubly Aligned Incomplete Multi-view Clustering (DAIMC)
      • Efficient and Effective Incomplete Multi-view Clustering (EE-IMVC)
      • Incomplete Multiview Spectral Clustering With Adaptive Graph Learning (IMSCAGL)
      • Self-representation Subspace Clustering for Incomplete Multi-view Data (IMSR)
      • Integrate Any Omics (IntegrAO)
      • Late Fusion Incomplete Multi-View Clustering (LF-IMVC)
      • Multiple Kernel K-Means with Incomplete Kernels (MKKM-IK)
      • Multi Omic Clustering by Non-Exhaustive Types (MONET)
      • Multi-Reconstruction Graph Convolutional Network (MRGCN)
      • NEighborhood based Multi-Omics clustering (NEMO)
      • Online Multi-View Clustering (OMVC)
      • One-Pass Incomplete Multi-View Clustering (OPIMC)
      • One-Stage Incomplete Multi-View Clustering via Late Fusion (OS-LF-IMVC)
      • Projective Incomplete Multi-View Clustering (PIMVC)
      • Scalable Incomplete Multiview Clustering with Adaptive Data Completion (SIMC-ADC)
      • Subtyping Tool for Multi-Omic Data (SUMO)
    • Decomposition
      • Data Fusion by Matrix Factorization (DFMF)
      • Joint Non-Negative Matrix Factorization (JNMF)
      • Multi-Omics Factor Analysis (MOFA)
    • Explore
      • Get summary
      • Get number of modalities
      • Get number of samples by modalities
      • Get complete samples
      • Get incomplete samples
      • Get samples
      • Get samples by modality
      • Get missing samples by modality
      • Get number of complete samples
      • Get number of incomplete samples
      • Get percentage of complete samples
      • Get percentage of incomplete samples
    • Feature selection
      • JNMF Feature Selector
    • Impute
      • DFMF Imputer
      • MOFA Imputer
      • JNMF Imputer
      • Missing Modality Indicator
      • Observed Modality Indicator
    • Load
      • M3Care Dataset Loader
      • MRGCN Dataset Loader
      • MUSE Dataset Loader
      • RAGPT Dataset Loader
    • Model selection
      • Multi-Modal Splitter
      • Train-Test Multi-Modal Split
    • Preprocessing
      • Drop Modality
      • Concatenate Modalities
      • Single Modality
      • Add Missing Modalities
      • Sort Data
      • Multi-Modal Transformer
      • Normalizer NaN
      • Select Complete Samples
      • Select Incomplete Samples
    • Retrieve
      • Multi-Channel Retriever (MCR)
    • Statistics
      • PID
    • Utils
      • Convert dataset format
      • Check Xs and y
    • Visualize
      • Plot combinations
      • Plot missing modality
      • Plot PID
      • Plot summary

Development

  • Contributing to iMML
    • Submitting a bug report or a feature request
      • How to make a good bug report
    • Contributing code
      • Pull Request Checklist
    • Guidelines
      • Coding Guidelines
      • Docstring Guidelines
    • API of iMML Objects
      • Estimators
        • Instantiation
        • Fitting
        • Transformers
        • Predictors
      • Deep Learning
        • Dataset defition
        • Estimator defition
        • Training
  • Project roadmap
    • Guiding principles
    • Thematic tracks
      • Expanding the multi-modal learning field
      • Algorithm diversity
      • Performance and scalability
      • Interoperability and compatibility
      • Documentation, tutorials, and examples
      • Community and governance
    • How to get involved
    • Notes
  • Changelog
    • Version 0.3.1
      • imml.classify
      • imml.cluster
      • imml.model_selection
      • imml.preprocessing
    • Version 0.3.0
      • imml.impute
      • imml.model_selection
      • imml.preprocessing
      • imml.statistics
      • imml.utils
      • imml.visualize
    • Version 0.2.0
      • imml.ampute
      • imml.classify
      • imml.impute
      • imml.load
      • imml.utils
    • Version 0.1.1
      • .github/workflows/ci_test.yml
    • Version 0.1.0
  • License

Links

  • GitHub
  • PyPI
imml
  • Search

© Copyright 2025, .

Built with Sphinx using a theme provided by Read the Docs.