Potential issues with large drift rates

[telegraphic]

We get a lot of hits at high drift rates, and I'm not sure if this is due to preprocessing artifacts, RFI, or bugs. Here's a histogram, taken from some test data:

Also wondering if this is due to some oddness I'm seeing with iterative blanking, where the number of hits increases after a blanking round:

[find_et]       Preprocess flagged:    1.27%
[find_et] (iteration 1 / 8) GulpPipeline.run: New hits: 22
[find_et] (iteration 2 / 8) GulpPipeline.run: New hits: 5
[find_et] (iteration 3 / 8) GulpPipeline.run: New hits: 3
[find_et] (iteration 4 / 8) GulpPipeline.run: New hits: 3
[find_et] (iteration 5 / 8) GulpPipeline.run: New hits: 2
[find_et] (iteration 6 / 8) GulpPipeline.run: New hits: 2
[find_et] (iteration 7 / 8) GulpPipeline.run: New hits: 3
[find_et] (iteration 8 / 8) GulpPipeline.run: New hits: 5

and

[find_et] (iteration 1 / 8) GulpPipeline.run: New hits: 14
[find_et] (iteration 2 / 8) GulpPipeline.run: New hits: 4
[find_et] (iteration 3 / 8) GulpPipeline.run: New hits: 6
[find_et] (iteration 4 / 8) GulpPipeline.run: New hits: 7
[find_et] (iteration 5 / 8) GulpPipeline.run: New hits: 6
[find_et] (iteration 6 / 8) GulpPipeline.run: New hits: 7
[find_et] (iteration 7 / 8) GulpPipeline.run: New hits: 6
[find_et] (iteration 8 / 8) GulpPipeline.run: New hits: 5

[telegraphic]

This just seems to indicate that wideband signals get picked up better at high doppler drifts, as the smearing correction averages and increases S/N. The iterative blanking has been fixed with the addition of signal extent calculation