CPT Hinge not honored

Q&A
1

Hello, I plot my Greenland data with polar cpt, my grid z values range from -0.8 to +0.2, so I create cpt as following:

makecpt -Cpolar.cpt+h -Ic -T-0.8/0.2/0.02

I was expecting the color changes from red to blue at 0, however, the resulting image is following:

GRACE_GFO_MEAN4_FM_EC_GRID_Trend.txt
GRACE_GFO_MEAN4_FM_EC_GRID_Trend.txt1577×1905 391 KB

Looks like the hinge at 0 is not honored. I am using GMT v6.0.0 on Windows.

BTW, I have to specify -Cpolar.cpt+h explicitly, if use -Cpolar+h, an error occured:
makecpt [ERROR]: File c:/programs/gmt6/share/cpt/polar+h.cpt not found

Thanks.

2

Hm, looks like we have a regression for those hinges. I will have a look.

3

Wait a minute. Running this works for me in 6.1 but I get the same as you in 6.0. I recommend you upgrade.

gmt makecpt -Cpolar+h  -T-0.8/0.2
-0.8	blue	0	white	L	
6.95312135747e-310	white	0.2	red	B	
B	black
F	white
N	127.5
4

Thanks. I just tried on our Linux server which has GMT v6.1.0 installed, however, the result is still exactly the same.

image
image800×131 16.9 KB

5

Are you saying that the command I showed above returns something else for you in Linux? Notice this is a continuous-color table so I would not add /0.02 unless you really want the colors to be constant within each 0.02 unit section.

6

I just tried to remove the /0.02 option from makecpt inputs, and then the result is correct, both on GMT v6.0 and v6.1. If I add /0.02 back, then the problem is back, both for v6.0 and v6.1.

Therefore, it’s the /0.02 option causing the problem, and this may be a bug in GMT?

7

Well, more like a feature. These are dynamic CPTs but when you start chopping them into constant slices I think the hinge stuff is deactivated.

8

Thanks, I will look into the concept of dynamic cpts etc.