Simulated-data

Use this repo to be used to generate simulated datasets for all kind of purposes within Epicentre (training, interviews, case-study ect...)

Measles Outbreak in Moïssala, Chad

This datasets (~ 5000 cases) simulates an outbreak of Measles across Southern Chad. It is made of a main linelist and a corresponding laboratory dataset. The linelist is built so that vacc_status, age_group, muac_cat are associated with a dead outcome.

Look at the .Rmd document for some ideas of analysis that can be done.

To replicate the data:

1. measles_data.R creates a lists of different paramaters and distributions for a measle outbreak, using the Chad data from the East Africa Surveillance dashboard

2. gen_linelist.R uses {simulist} to generate a realistic linelist for the Measle outbreak. It then attributes symptoms and outcome based on a logistic regression model. 3.gen_linelist_geo.R uses the simulated linelist and adds geographic variable (Region, Sub-prefecture, Villages and Health facility) in a way that ressemble a spatially aggregated outbreak.

3. gen_lab_data.R uses the clean simulated linelist to generate a laboratory datasets of confirmed cases.

4. make_raw_linelist.R takes both the simulated linelist and the laboratory data to add some dirtiness and change the columns names. The data are then exported to .xlsx and .csv. A variable dictionnary summarising clean/raw data names, categorical values and number of missing values is also created and saved in data/dictionnary/

5. linelist_fr_translation translates both the clean and the raw data to french (variable names and categorical values), and creates the relevant variable dictionnary.

6. Because these data are used in the {repicentre} repository, you can run copy_data_to_repicentre.R to copy the latest data in the repository.