Feature #516
Should there be a makecpt -E<nlevels> option?
| Status: | Closed | Start date: | 2014-02-13 | |
|---|---|---|---|---|
| Priority: | Normal | Due date: | ||
| Assignee: |  Paul | % Done: | 100% | |
| Category: | - | |||
| Target version: | Candidate for next minor release | |||
| Platform: |
Description
grdinfo -E<nlevels> reads the range of a grid, splits it into <nlevels> intercals and builds a cpt. It allows a quick cpt suitable for a grid to be generated. Yet, there is no direct equivalent for table data. You can use gmt info -T<zinc> to return a -Tzmin/zmax/zinc string that can be used with makecpt to create the cpt file, but this result depends on <zinc> which would have to be selected carefully, knowing the data range in the first place. Experienced scripters can deal with this in a few lines, but it is certainly more tedious than for grids.
Proposal: Add a new -E<nlevels> option to makecpt. This would imply two things:
- A filename(s) (or stdin) will be required so makecpt can compute the range of the data set.
- Once range is known then the cpt file with <nlevels> slices can be computed.
Users would use the common -i<cols> option to select from which data column the range should be computed. If not given we default to the z column [-i2] or the last column, whichever is first.
Associated revisions
Initial implementation of issue #516 for makecpt new -E option
History
#1
Updated by Florian over 3 years ago
grdinfo -E<nlevels>
You mean grd2cpt, right? I like your proposed makecpt feature. The resulting cpt file should ofc be histogram-equalized. So apart from determining the range, makecpt also has to fit the levels such that the number of records per level is equal.
#2
Updated by Paul over 3 years ago
- Status changed from New to Feedback
Sorry, yes grd2cpt obviously. Histogram equalization is a different matter. -E<nlevels> in grd2cpt is just a simple stretching, there is no equalization there; that takes place via other options or no option.
#3
Updated by Remko over 3 years ago
Yes, I like the either of added the option -E also to makecpt.
Can I ask for another feature to -E?
It would be nice if -E without any argument would create the same amount of levels as the colour map contains. This would be similar to makecpt -T<zmin>/<zmax> without specifying the interval. Of course, both makecpt -E and grd2cpt -E should have this.
#4
Updated by Paul over 3 years ago
Sure, I agree it should be -E[<nlevels>], with the default <nlevels> being picked from by the cpt file.
#5
Updated by Paul over 3 years ago
- Status changed from Feedback to Resolved
- % Done changed from 0 to 100
I have made the initial implementation of -E as discussed; in r12909. The last data column is used for the range determination and this can be changed via -i. For native binary data tables the -b option must be used.
#6
Updated by Paul about 3 years ago
- Status changed from Resolved to Closed
Closing as it is implemented.