create data layer for worldpop_age_sex

[tedw0ng]

Create data layer for Worldpop disaggregated data by age and sex groups.
Question: the regular worldpop layer has a .mean() line. Should it be sum()? Same question would apply to this new layer.

[S-AI-F]

@tedw0ng Agree that it should be sum() instead of mean(). I checked the original notebook and we are using also mean(): https://github.com/wri/cities-indicators/blob/emackres-patch-1/notebooks/extract-layers/extract-WorldPop-OpenSpaceTreeAccess.ipynb
I don't remember any reason for that. I think we should change this!

[S-AI-F]

I also see that worldpop provides the data through a specific api now: https://www.worldpop.org/sdi/introapi/
Should we consider changing the data access?

[tedw0ng]

I think the Worldpop API only returns JSON data, not rasters. That might be okay for indicators, but not for data layers.

[S-AI-F]

It actually serves also the url to get the tiff files but at the national level. So it is probably better if we keep extracting the data from GEE to get only areas of interest

[weiqi-tori]

Could add a test for the layer to check if it works well

[weiqi-tori]

agesex_classes should be added as an input parameter for __init__()

[tedw0ng]

added agesex_class as input parameter

[weiqi-tori]

The layer and test standard structure changed since this layer was created, so I fixed the conflict and updated the layer.

@tedw0ng: Please pull the branch with the latest update, and test if the layer functions as you expected. One note is that I switched the aggregation for image collection back to mean(). It's not aggregate the pixels within the bbox, it's for the image collection aggregating to an image when there is overlap between images.

[S-AI-F]

@weiqi-tori @chrowe

Checked in this notebook and works globally well: https://colab.research.google.com/drive/1VtDgtQ_BVRh7h0QJ56ryfHJCByLs5Rqu#scrollTo=uXq9pqNOH7fw

Proposed changes:

• Would be great if we add an error checking on the "year" parameter. It should be 2020 and if the user specify different year, show an error message: "the selected year is not available". Now it works with any year value
• Can we add in the code the available age_sex list of values: [M_0, M_1, M_5, M_10, M_15, M_20, M_25, M_30, M_35, M_40, M_45, M_50, M_55, M_60, M_65, M_70, M_75, M_80, F_0, F_1, F_5, F_10, F_15, F_20, F_25, F_30, F_35, F_40, F_45, F_50, F_55, F_60, F_65, F_70, F_75, F_80]

[S-AI-F]

add comment n possible list of values here just for documentation: [M_0, M_1, M_5, M_10, M_15, M_20, M_25, M_30, M_35, M_40, M_45, M_50, M_55, M_60, M_65, M_70, M_75, M_80, F_0, F_1, F_5, F_10, F_15, F_20, F_25, F_30, F_35, F_40, F_45, F_50, F_55, F_60, F_65, F_70, F_75, F_80]

[S-AI-F]

@weiqi-tori url for dataset documentation in airtable: https://airtable.com/appDWCVIQlVnLLaW2/tblYpXsxxuaOk3PaZ/viwExxAgTQKZnRfWU/recFjH7WngjltFMGi?blocks=hide