The guess cadence routine works quite reliably, so much that it's confusing when it doesn't work. This ticket is to collect cases where the cadence fails.
I put in code to check that the cadence guessed won't break more than 25% of the connection, when there are fewer than 65 points. This was motivated by /home/jbf/ct/autoplot/script/development/interpolate/cadence/cadenceFailRank1Small.jy
Jon N at UNH had a case where the highest energy of a spectrogram was cut off from the rest. See sftp://jfaden.net/home/jbf/ct/autoplot/u/2022/jonn/20220223/het_dth_comparison.vap
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I need to start an area of data/vaps which have cadence breaks when they shouldn't, and when they don't but should. See sftp://jfaden.net/home/jbf/ct/autoplot/cadence/examples.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
A single-element dataset passed into guessCadence would result in a IndexOutOfBoundsException. This case is handled now. I need to remind myself where all the corner cases are kept and add the single-point dataset to the cases.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The data cadence slows down when the speed slows down:
vap+jyds:https://jfaden.net/~jbf/autoplot/data/kmz/Jul_8,_2024_12_58_54_PM.kmz/doc.kml?speed&script=https://github.com/autoplot/dev/blob/master/demos/formats/kml/tracksKML.jyds
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
This part of an orbit successfully guesses a cadence, when there should not be.
Last edit: Jeremy Faden 2016-05-11
The problem there was that there was not enough data to do good statistics. In this case it should just return the highest bin where data was seen.
http://jfaden.net/jenkins/job/autoplot-test014/ is the grand test of cadence calculation.
Last edit: Jeremy Faden 2018-08-07
I've also added http://apps-pw.physics.uiowa.edu/hudson/job/autoplot-test070/ 014.
I put in code to check that the cadence guessed won't break more than 25% of the connection, when there are fewer than 65 points. This was motivated by /home/jbf/ct/autoplot/script/development/interpolate/cadence/cadenceFailRank1Small.jy
Last edit: Jeremy Faden 2016-06-23
Here's a simple 9-point set.
Ivar has a guess-cadence failure here: nudnik:/home/jbf/project/rbsp/users/ivar/20170510.
See sftp://jbf@jfaden.net/home/jbf/ct/autoplot/bugs/1582/MMS_FPI.vap, which fails to guess cadence for some intervals.
Last edit: Jeremy Faden 2017-07-23
Ivar pointed out a couple of new guess-cadence failures. Also he requested that the spectrograms mind the "cadence check" property.
Note the series renderer is suddenly ignoring the cadenceCheck property.
Last edit: Jeremy Faden 2018-10-09
See also https://sourceforge.net/p/autoplot/feature-requests/652/
The series renderer is again respecting the cadenceCheck property.
https://github.com/autoplot/dev/blob/master/bugs/2022/20220125/shouldNotBreak.jy
See https://github.com/das-developers/das2java/issues/26 which might affect this ticket.
Jon N at UNH had a case where the highest energy of a spectrogram was cut off from the rest. See sftp://jfaden.net/home/jbf/ct/autoplot/u/2022/jonn/20220223/het_dth_comparison.vap
See also https://github.com/das-developers/das2java/issues/15.
This was causing new problems with test038, so I reverted the logic.
https://github.com/autoplot/dev/blob/master/bugs/2022/20220307/test038guessCadence.jy
Last edit: Jeremy Faden 2022-03-21
See https://sourceforge.net/p/autoplot/bugs/2440/ where a change for this caused new problems with test038.
It turns out there are a bunch of cases which were fixed by the new cadence logic, so I reverted the revert.
See https://github.com/autoplot/dev/blob/master/bugs/sf/1582/guessCadenceFailGCD.jy which shows where random spacings of 1,2 or 3 fail to identify the GCD of 1.
I need to start an area of data/vaps which have cadence breaks when they shouldn't, and when they don't but should. See sftp://jfaden.net/home/jbf/ct/autoplot/cadence/examples.
https://github.com/das-developers/das2java/issues/38
A single-element dataset passed into guessCadence would result in a IndexOutOfBoundsException. This case is handled now. I need to remind myself where all the corner cases are kept and add the single-point dataset to the cases.
The data cadence slows down when the speed slows down:
vap+jyds:https://jfaden.net/~jbf/autoplot/data/kmz/Jul_8,_2024_12_58_54_PM.kmz/doc.kml?speed&script=https://github.com/autoplot/dev/blob/master/demos/formats/kml/tracksKML.jyds
Bob pointed out where the output of FFT has a return which is drawn but shouldn't be:
https://sourceforge.net/p/autoplot/bugs/2684/
Last edit: Jeremy Faden 2025-03-04