Skip to content

Latest commit

 

History

History
28 lines (20 loc) · 1.26 KB

developers.md

File metadata and controls

28 lines (20 loc) · 1.26 KB

README for developers

WIP

Familiarize yourself with the structure of the tool

./AItomotools     : The toolbox
├── CTtools                 -> Tools for Computed Tomography specifically
├── data_loaders            -> Tools for 1) preprocessing data 2) loading data for training
├── metrics                 -> Tools to evaluate performance
├── models                  -> Machine Learning reconstruction models
└── utils                   -> Other type of generic utils

./demos           : Folder with demos to learn about the toolbox. Please add demos about your new functionality

./scripts         : Scripts that fully reproduce papers/data
├── data_generation_scripts -> Scripts that produce the datasets.
├── paper_scripts           -> Code that reproduces published papers

./utils           : Other utilities. e.g. pre-commit scripts. 

Aitomotools base class

Go to ./AItomotools/models/LIONmodel.py to see the base class. All models should inherit from this. It provides useful tools, like save() , load(), save_checkpoint(), default_parameters() etc. You must make your models use these functions because they save more than the models, also parameters and commit hashes, so the code is more reproducible.