birm / datadash Goto Github PK

1d

• QQ against normal
• hist
• bar

2d/3d

• qq between variables
• time series
• scatter

Maybe support deterministic random sampling given seed.
Something requested is to find a subset of a population matching a different population.
I’m not sure how I want to go about this, but some ideas:
Knn without replacement
a is set to match, b is set of candidates to use, b’ is selected set of b.
Start with b’ empty
For each in population a, find nearest in population b, add to b’, remove from candidate pool, repeat
May want to set a threshold over which to report seperately

• Table
• Image Grid
• SPLOM