[ncl-talk] Emphasizing statistically significant data in contour plots

WU, Cheuk Kuen hkson at connect.hku.hk
Wed Aug 6 12:00:44 MDT 2014


Hi,

I also have another question for my MJO study plots apart from the panel
plots. But it is quite different and I choose to start a new post.

I have done the t-test for my data to test their significance. I have also
calculated the critical t-values for each grid by using cdft_t function.
For vectors, I plot all the arrows in gray colour first. Then I do another
plot in which all insignificant data are set to missing values (by using
ndtooned, idx and onedtond). The arrows for significant data are black in
colour. The two plots are then overlaid. It is done well for the vector
plot.

For contour plot, like geopotential height data, I first plot all data with
cnFillOn=True and cnLinesOn=False. Again I do another plot and the
insignificant data become missing. The significant data are plotted with
both cnFillOn and cnLinesOn set to true, and the two plots are overlaid, so
that the significant data can be emphasized with the contour lines.
However, I realised that this should not be a good method because some data
WITHIN the contour fill may be significant but cannot be emphasized.

I think of using cnFillPattern. For example, I want there will be dots over
all the significant data. I have not yet tried this, but I wonder that
whether the fill patterns have background colour (e.g. white)? If yes, they
will cover my original cnFill after overlaying the plots. I want the fill
colours of all data can still be seen with dots over the significant data.
Is this method possible? Or do you have even better suggestions? Thank you!

John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140807/a8d96369/attachment.html 


More information about the ncl-talk mailing list