Welcome to stc_unicef_cpi’s documentation!

stc_unicef_cpi is a Python library for the open source data science for international development community that estimates high-resolution, multi-dimensional child poverty using DHS (and soon MICS!) geocoded survey data.

The package was initially developed during DSSGxUK 22, in collaboration with Save the Children and UNICEF.

Check out the Getting started section for further information about how to begin, including Installation of the package.

Raw data used is explained in Data, methodology in Methodology (with sections on e.g. Data preparation etc.), and some initial results are displayed in Results.

Information for contributors is available in Developer information.

Note

This project is under active development.

Contents

Table of Contents

• Home
• Getting started
  • Installation
  • Getting data
  • Training a model
  • Making predictions
• Reproducing results
  • Reproducing data
  • Reproducing models
  • Reproducing predictions
  • Additional steps
• Data
  • Demographic and Health Surveys (DHS)
  • Open Street Maps (OSM)
  • Google Earth Engine (GEE)
    • Night-time lights data
• Methodology
  • Data preparation
  • Evaluation
  • Label transformation
  • Explainable ML pipeline
• Results
  • Conclusions
  • Deliverables
  • Future work
• DL steps
• Developer information
  • Getting started
    • Install locally
  • TODO Options
    • assignees:
    • labels:
    • milestone:
    • Other Options
    • Reference
  • Project Organization
  • Script breakdown
  • Main components
  • Contribution guidelines
    • Developer information
• API
  • stc_unicef_cpi.data
  • stc_unicef_cpi.features
  • stc_unicef_cpi.models
  • stc_unicef_cpi.utils
  • stc_unicef_cpi.visualization

Microestimates of Child Poverty in Developing Countries

A package to produce high-resolution multi-dimensional child poverty index estimates.

Project Scope

Pass

References

References to be added here in future.

Indices and tables

• Index

• Module Index

• Search Page