databrewllc/anomaly-detection repository overview

⁠DataBrew ODK Anomaly Detection

This repository is used for DataBrew ODK forms anomaly detection. Survey forms will be processed through data cleaning process and anomaly identification. This process will going through continuous iteration to increase survey data quality.

⁠Reproducing Environment

⁠Clone repository

git clone https://github.com/databrew/anomaly-detection.git

This project uses renv to restore all the libraries being used for data processing. It is recommended to use renv library to reproduce this analysis

⁠Install Renv

install.packages("renv")

⁠Initiate Renv

library(renv)
renv::restore()

The above command will reproduce the analysis environment used for data stored in S3

⁠Contributing

To contribute, design your new feature in a new branch and create a PR to the main branch when ready

⁠Create new branch

git checkout -b [name_of_your_new_branch]
git push origin [name_of_your_new_branch]

⁠Data Cleaning Functions

All data cleaning functions are here⁠. Once function is created, append your function in here⁠

⁠Anomaly Detection Functions

All anomaly detection functions are here⁠. To add more anomaly identification procedure, create a new function that takes in registration/household data that returns a tibble dataframe with these 3 columns type, anomaly_id, description. Once function is created, append your function in here⁠

⁠Deploying your Changes

Github repository has been integrated with Github Actions, doing git push to this repo main branch will automatically sync with (DockerhHb)[https://hub.docker.com/r/databrewllc/anomaly-detection⁠]

⁠References

• Anomaly detection Google Sheets⁠