Introduction

This repository contains a collection of Python scripts and Jupyter notebooks to implement spatial multivariate imputation using Machine learning and Lambda distribution.

Python Setup

It is recommended to use Anaconda Python distribution and create an environment using the provided environment.yml file.

Notebooks

The notebooks use the Python scripts that are located in Tools folder.

• original.dat is the input data file.
• 01-DataInventory.ipynb the first notebook that takes care of data processing and exporting the data for imputation.
• 02-LambdaDistributionMlModel.ipynb this notebook is used to fit the lambda distribution and its results will be used by downstream notebooks.
• 03-MlForConditionalDistributionTemplate.ipynb a template notebook to implement training of the neural networks. It is not required to run this notebook as it is used by the next notebook to execute training.
• 04-DataImputationUsingMLP.ipynb the main notebook that implements all the steps of imputation and cross validation. This notebook uses papermill python package to orchestrate training of neural networks.