RobHolman
commented
3 years ago Obviously I am a big fan of this update. Thanks for Jed for loading it up. Here are a few comments as they occur to me:
- As mentioned on email, the chosen demo run, August 8, 2015, is pretty anomalous and would not be a great demo run (> 1.2 m Hs). We could choose a better run.
- For the demo results jpg, I think we should plot both the bathy and the bathy error (hErr) using two subplots. In fact, it is best to just use plotBathyCollect(bathy) to automatically do this. People need to be more aware of hErr and not just assume all bathy results are equal.
- As a GitDunce I’m not sure where the user manual is, but I’m guessing that it is by now ancient and should be updated. (Ah, it is in the wiki – should it also be a pdf?)
- In the demo, why not use plotBathyCollect(bathy) to do the final plotting
- I know that some of the routines I use take advantage of CIL conventions, but I love the discipline that they impose. For instance, argusFilename and parseFilename are pretty elegant compared to doing things longhand. I bet they could be made more generic pretty easily.
- There is a pdf cBathyVersionCompare etc that describes at length the previous versions and what is new in V2.0. Should this be in the wiki or somewhere obvious?
- The folder labelled “manual” doesn’t have the manual, just some png updates (I think).
- The README notes that fixBathyTide is a CIL routine. I believe that it tries to execute whatever tide function the user specifies in params.tideFunction. The example says cBathyTide, which uses our database, but other users should be aware that they need supply a working tide routine. I think this is described in the user manual (wiki), but perhaps we need to have a more specific description.
- Have we included the ground truth data, particularly bathyTestDataSubSet which includes all the survey and environmental info??
- I’m struggling with the README end section of logical flow. We made up some figures and tables for this that might help.
- I feel like we have a lot of files that are obsolete, like anything with aerielle in it, democBathyUAV, findKAlphaSeedObs, loadBathyStack, results_cBathy_dwell_test etc, testStack102210Duck.
- Is driver_cbathy_Feb_2019 something that should be in the toolbox?
- I note that Jed has included a lot of code that goes against normal argus conventions, particularly variable naming (we use camelback or whatever it is called) and cBathy has a cap- B
- Logic.txt likely needs updating and should be part of the updated README
eswanson-usgs
The upgrade to Version 2.0 involves three significant changes, so is considered a major revision and is given a new leading version number. A brief description of the changes is given below, and a detailed comparison of cBathy versions is provided in a PDF file (cBathyVersionComparison_Dec14_2020.pdf) included in the main directory of the repository.
In order of importance, the changes are a) to change from tile sizes that are fixed by the user to those that change depending on expected wavelengths, b) to change from solving for three variables, k, alpha, and phi, at each map point to solving for only k and alpha, and c) introduction of a better algorithm to find seed values for k and alpha before the nonlinear search for each tile.
This version was tested using the function 'democBathyVersion2p0.m', which runs the main cBathy processing to create phase 1 and 2 results (fDependent and fCombined) and plots the bathymetry results from fCombined. It is currently set to analyze a test data set, 1439053140.Sat.Aug.08_16_59_00.GMT.2015.argus02b.cx.mBW.mat, one of the new standard testbed data sets created by OSU. The cBathy input data is contained in the subfolder DemoData along with the corresponding bathymetry output. The demo produces and saves the bathymetry information estimated by cBathy into a new folder.