some miscellaneous updates #225

WPringle · 2021-05-13T03:04:40Z

change order in cmocean for applying the layer interpolation to the end (seems better to me)
change order of applying cmocean in plotter to the end
adding radius_separated_points function that is to be used before m_quiver function to ensure that you have get equal spatial distribution of arrows
change to tidal_data_to_ob function so that it keeps all tidal constituents for the boundary but sets to zero those which are not in the database to make it easier for user to set themselves later if desired.
correcting m_grid call in msh.plot() to use the background color (backcolor option)

…he end (seems better to me) - change order of applying `cmocean` in `plotter` to the end - adding `radius_separated_points` function that is to be used before m_quiver function to ensure that you have get equal spatial distribution of arrows - change to `tidal_data_to_ob` function so that it keeps all tidal constituents for the boundary but sets to zero those which are not in the database to make it easier for user to set themselves later if desired.

krober10nd · 2021-05-13T03:25:47Z

utilities/radius_separated_points.m

+%  so that they are at least separated by a distance of
+%  user-defined 'radius'
+
+% computing rangesearch


Please use ourRangeSearch to avoid toolbox dependency if possible

ourRangeSearch doesn't seem to work for multiple points.

you call it in a loop over all the query points.

see this line there's some unfortunately hardcoded parameters.
[idx,~] = frsearch(anno, testset, r, 1e3,0,1);

ok updating that now..

latest commit addresses this

but note for test I was using it took 3 sec for rangesearch but 50 sec for ourRangeSearch.

yea, it's not vectorized. I guess you use this for quiver plotting. What you could do is check if the rangesearch exists if they have the toolbox and then default to this.

yeah it does do that

krober10nd · 2021-05-13T03:27:38Z

utilities/tidal_data_to_ob.m

-       keep(j) = false;  
-       continue
+        disp(['No tidal data in file for constituent ' const{j}])
+        obj.f15.opealpha(j).name = const{j};


What’s the use case for this? I didn’t understand your explanation in the commit

The use case I was doing is the following:
const = ["Z0" "major8"];

The Z0 is the zero frequency constituent that is actually valid in the constituent list but TPXO8 provides no data. So before I would just have this deleted from the boundary condition list. However, I was using Z0 to impose a sea level rise offset.

So I was using Make_f15 to populate Z0 with "zeros" as a dummy into the boundary constituent list then overwrite it later with the desired amplitude. An additional thing I was thinking to add was in writefort15.m that it would only write it out if it wasn't all zeros to avoid cluttering of fort.15 but I think this could be considered a fairly minor concern unless you have many constituents.

so in the case you're applying a SLR offset using this tidal constituent, you also add a geoid offset to the NA file?

no I did not but I suppose you could (should?) do that as initial condition yes

I was using small domain and just spun it up slowly to that state.

yea I think you should specify an initial condition that matches the elevation of the boundary condition otherwise you'll create a large gradient near the boundary in free surface elevation.

but this for some more advanced users anyway.

nah but since you spin it up from zero it works. And also even if elevation is below the new sea level with SLR it may not get flooded due to connectivity so that's why I just did the boundary condition

…oints could end up going slightly below this causing NaNs in the projection

…ts themselves

…ints are not pushed outside and become NaNs (was limited to radius of 178 deg but made sure can go up to full 180 deg)

krober10nd · 2021-05-30T15:05:01Z

change order in cmocean for applying the layer interpolation to the end (seems better to me)

What does this mean?

WPringle · 2021-05-31T16:33:08Z

change order in cmocean for applying the layer interpolation to the end (seems better to me)

What does this mean?

So in cmocean it was applying the interpolation of the continuous colormap to discrete levels somewhere in the middle of the code, however I changed it to the end.

The reason being was that for the colormap with pivots the interpolation to the discrete levels before the pivot application was causing problems because the pivot will not be precisely applied once the colormap has been reduced to a few discrete levels. I found better performance with it at the end with pivots e.g., when plotting the bathymetry 'b'.

krober10nd · 2021-05-31T17:02:08Z

yea, there's often random issues with plotting some quantities that require you tweak the colormap option. I guess this helps reduce the occurrence of those issues.

…to ourRangeSearch in case rangesearch does not exist in radius_separated_points function

krober10nd · 2021-05-31T17:44:33Z

changelog?

WPringle · 2021-05-31T17:57:36Z

changelog?

look ok?

krober10nd · 2021-05-31T17:59:19Z

Looks great to me.

krober10nd reviewed May 13, 2021

View reviewed changes

WPringle added 3 commits May 18, 2021 16:46

correction to bbox for global mesh - bottom was set at 88S and some p…

c74397a

…oints could end up going slightly below this causing NaNs in the projection

ignoring all files in Examples directory except for the example scrip…

4ded5d9

…ts themselves

correction in setting stereographic projection bounds to make sure po…

4ad33f7

…ints are not pushed outside and become NaNs (was limited to radius of 178 deg but made sure can go up to full 180 deg)

correcting ourRangeSearch for multiple query points, and adding call …

e72cc42

…to ourRangeSearch in case rangesearch does not exist in radius_separated_points function

krober10nd approved these changes May 31, 2021

View reviewed changes

populating changelog for pull 225

31d3876

WPringle merged commit 543ffb6 into Projection May 31, 2021

WPringle deleted the misc_updates branch May 31, 2021 17:59

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

some miscellaneous updates #225

some miscellaneous updates #225

WPringle commented May 13, 2021 •

edited

Loading

krober10nd May 13, 2021

WPringle May 31, 2021

krober10nd May 31, 2021

krober10nd May 31, 2021

WPringle May 31, 2021

WPringle May 31, 2021

WPringle May 31, 2021

krober10nd May 31, 2021

WPringle May 31, 2021

krober10nd May 13, 2021 •

edited

Loading

WPringle May 31, 2021

krober10nd May 31, 2021

WPringle May 31, 2021

WPringle May 31, 2021

krober10nd May 31, 2021

WPringle May 31, 2021

krober10nd commented May 30, 2021

WPringle commented May 31, 2021

krober10nd commented May 31, 2021

krober10nd commented May 31, 2021

WPringle commented May 31, 2021

krober10nd commented May 31, 2021

some miscellaneous updates #225

some miscellaneous updates #225

Conversation

WPringle commented May 13, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

krober10nd May 13, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

krober10nd commented May 30, 2021

WPringle commented May 31, 2021

krober10nd commented May 31, 2021

krober10nd commented May 31, 2021

WPringle commented May 31, 2021

krober10nd commented May 31, 2021

WPringle commented May 13, 2021 •

edited

Loading

krober10nd May 13, 2021 •

edited

Loading