This pull request fixes a very small formatting issue in a doc, then adds a new testcase for the LRO NAC linescan sensor. (The location is somewhere in the Shackleton crater, at the South Pole.)
This testcase was giving the linescan sensor ground-to-image operation a very hard time before the current secant method got implemented, because the initial linear approximation was off by up to 100 pixels (when the ground is 2000 meters above the nominal datum). I am hoping this can be the go-to dataset whenever we try any improvements to the linescan sensor.
While one may think that this dataset is a little big, at 511K, it is a real-world dataset, and current toy test data can help only so much in ensuring the quality of a package like this which will be used for anything under the Sun (within the Solar System, that is). Also, the json package which this depends on has a jeopardy.json testcase which is 53 MB.
This pull request fixes a very small formatting issue in a doc, then adds a new testcase for the LRO NAC linescan sensor. (The location is somewhere in the Shackleton crater, at the South Pole.)
This testcase was giving the linescan sensor ground-to-image operation a very hard time before the current secant method got implemented, because the initial linear approximation was off by up to 100 pixels (when the ground is 2000 meters above the nominal datum). I am hoping this can be the go-to dataset whenever we try any improvements to the linescan sensor.
While one may think that this dataset is a little big, at 511K, it is a real-world dataset, and current toy test data can help only so much in ensuring the quality of a package like this which will be used for anything under the Sun (within the Solar System, that is). Also, the json package which this depends on has a jeopardy.json testcase which is 53 MB.
This is also relevant to issue https://github.com/USGS-Astrogeology/usgscsm/issues/360.