Bug roll up for Flux file
- SchattenErrorModel and SchattenTimingModel give erroneous results for some configs.
Status Fixed in 9-21 build.
- Need to define file requirements. Header is not clear and system is quite brittle to minor changes and crashes regularly if header is not exactly like the test case file header. All files that have "Crash" in the schatten file name currently cause a crash.
Status: The file read portion was rewritten. None of the "crashes" inputs crash on Linux now. Needs to be checked on Windows.
- Need to implement interpolation of F107 values. Newly created stress tests shows major different interpolation can have.
- Requesting epoch before beginning of file throws. See script named SchattenPredict_PropOntoBeginningOfFile
SolarSystem exception: "SolarFluxReader::GetInputs()" Index can not be less than zero.
- Requesting epoch after end of file yields NaN force model evaluation. See script named SchattenPredict_PropOffEndOfFile
- Previous propagations using different spacecraft and different propapagator affect propagation with schatten file. Run Schatten_Wierd. If you comment out the first propagate command, the results from the second command are different in the report.
Fixed. Need to watch test results for potential side effects.
- Startup file points to a default atmosphere file. If that file is not present, then all tests fail. Note sure why we need that file, is it for the default? Need to be careful here and make sure we are only using that file when needed.
Needs final confirmation after file validation is improved, it appears these issues we're due to invalid files.
- Crash on re-run. Load a script from the toolbar that uses a schatten file (attached scripts should duplicated). Run the script using Play button in the toolbar. After run is complete, in the resource tree "Script" folder right click on the script and select "Save-Sync-Run". I see a crash occur most of the time. Sometimes I have to Save-Sync-Run multiple times. (found in Sept. 18 nightly build)
This appears to have been a character encoding issue. I didn't see it on Linux, but did on Windows due to line endings. The line by line read was then fragile. The reader has been rewritten to fix it.
- Propagating using a Schatten file in a For loop gives erronouse results the first time. Re-running from the "play" button avoids crash above and gives correct results. See script named SchattenForLoop.script