As a good way to know that data are converging towards their "best" fit, it's important to visualize how the fitting coefficients change on each iteration.
global vs group vs individual
individual stations -> "while there's nothing left to throw out, keep iteratively fitting and tweaking coeffs"
This can be expected to be much more intensive, as the data for an individual cast should be plotted against one another and all the coefficients (with matching bottle and CTD flags) recorded into one big happy report that a human can read. This type of logging for iterative fitting will be quintessential for highlighting different fitting techniques and convincing PIs that the "algorithms are good". It's also good for instructional purposes.
[ ] Add CLI or SSSCC option to fit a cast "verbose"
[ ] For each iteration of fitting done on that selected cast, generate a plot of the profile
[ ] Record all the fitting coefficients
[ ] Record the fit statistics (if possible, give individual RMS errors)
[ ] Record what the flags would be if using the current coeffs
As a good way to know that data are converging towards their "best" fit, it's important to visualize how the fitting coefficients change on each iteration.
This can be expected to be much more intensive, as the data for an individual cast should be plotted against one another and all the coefficients (with matching bottle and CTD flags) recorded into one big happy report that a human can read. This type of logging for iterative fitting will be quintessential for highlighting different fitting techniques and convincing PIs that the "algorithms are good". It's also good for instructional purposes.