Sara Stoudt
Sara Stoudt
expect to be replaced by NA?
update: I don't think what I initially proposed is necessary anymore. The headers for the lowest and highest bins are fine. A user could optionally provide a jam value for...
Proof of concept for sums [here](https://github.com/sastoudt/census-data-aggregator/blob/disaggregate/disaggregation.ipynb). Means is still in progress, was only accounting for margin of error in `n`, not `val`.
Means proof of concept done at above link. This approach will need some specific tests/checks and additional inputs from the current census aggregator outputs.
Medians proof of concept done at above link. Now needs a plan for integration with census aggregator and checks/tests
Added warnings for some checks and some documentation
addressed in https://github.com/datadesk/census-data-aggregator/pull/20
helper function [here](https://github.com/sastoudt/census-data-aggregator/blob/diagnostics/diagnostics.ipynb)
Proof of concept [here](https://github.com/sastoudt/census-data-aggregator/blob/varReplicates/varianceReplicates.ipynb) for sums, means, and medians. This assumes we have the dataset that contains the variance replicates. Grabbing these more automatically might require a different tool. This...
Proof of concept updated into function form with documentation.