dhs_regions

[GIS243]

Hi

I hope you are doing well.

Can you please let me know how you are getting the dhs_regions.csv file?

Thanks

[ardieorden]

Hi @GIS243 , the dhs_regions.csv file was created using one hot encoding. The categorical variables (i.e. region name) were converted into a numerical data that was either 1 or 0.

For example, if cluster 1 is in Region 1 and cluster 2 is not in Region 1, then the CSV will look like this:

dhs_cluster,Region1,Region2,...
1,1,0,...
2,0,0,...

[GIS243]

Thanks

can you please share a code for this thanks

[ardieorden]

Hi @GIS243 , I'm not sure if our code for doing this is currently available in the repo.

But here are some code samples that you might be able to use:

• https://machinelearningmastery.com/how-to-one-hot-encode-sequence-data-in-python/
• https://www.geeksforgeeks.org/ml-one-hot-encoding-of-datasets-in-python/

[RizwanZa]

Hi

I hope you are doing well.

Can you please send me dhs_indicators.csv file or .DTA and .DO?

[ardieorden]

Hi @RizwanZa ,

Unfortunately, due to data access agreements, we cannot send the .dta and .do files.

You would have to download data these files yourselves from the DHS Program website (instructions) and then you can create the dhs_indiators.csv file by running the /notebooks/00_dhs_prep.ipynb Jupyter notebook.

[jtmiclat]

Hi! Closing this issue as ill be archiving this repository. To see our new poverty mapping project checkout https://github.com/thinkingmachines/unicef-ai4d-poverty-mapping