Re: making dots in pattern 17 smaller

Hi Rick,
sorry for the delay and to be honest I've only made a small amount of progress.

I was using panelling and this may have an impact on dot size. Recently to get a desirable result I was requesting more panels than I actually used (to avoid an issue of parts of the image falling off the page).

I have uploaded 2 figures, 1.pdf and 2.pdf to your site.

The first one uses and reserves a single panel. The second one uses 1 but reserves 5 panels. We can see that clearly the dot size and density changes, I assume this is b/c of a primitive graphics feature rather than a bug?

I will probably revisit my panelling approach, as this will help to keep the dots to a good small size. I might have a question about this later.

Many thanks,
Mark.

On 02/10/2012, at 11:35 AM, brownrig@ucar.edu wrote:

> I tried adjusting the limits in NGDOTS, without apparent effect. The
> dots in the coneff_10.ncl are already so fine, I think we've reached a
> fundamental limit as to size vs. visibility. None of this explains
> why the dots in Mark's example are relatively large. Mark -- I'd be
> happy to take a look at your script, if you care to send it.
>
> RB
>
> On Mon, 1 Oct 2012 15:00:25 -0600
> David Brown <dbrown@ucar.edu> wrote:
>> I just examined the NGDOTS code (which is used to actually draw each
>> dot for fill pattern 17) and found 1) that even though, as Rick
>> noted, the minimum value that can be specified for fill dot size is
>> 0.0001, in fact it looks like NGDOTS does not permit a size less than
>> 0.001. The limitation appears designed to prevent the dots from
>> becoming so small that some of them disappear completely. I agree
>> with Rick that Mark's dots look bigger that what is possible even
>> with these limitations in NCL. But perhaps it is worth experimenting
>> with adjusting the limits in NGDOTS.
>> -dave
>>
>>
>> On Oct 1, 2012, at 2:07 PM, Rick Brownrigg wrote:
>>
>>> Hi Mark, Adam,
>>>
>>> There appears to be an internal limit of 0.0001. However, I set it
>>> to a smaller value and it did not have any effect, so I suspect
>>> there's other limits coming in to play.
>>>
>>> Mark, I'm surprised at how large the dots are in your plot; example
>>> 10 in the link achieves much finer dots, and illustrates different
>>> spacings.
>>>
>>> http://www.ncl.ucar.edu/Applications/coneff.shtml
>>>
>>> I don't know what else to suggest at this point…
>>> Rick
>>>
>>> On Oct 1, 2012, at 11:57 AM, Adam Phillips wrote:
>>>
>>>> Hi Mark,
>>>> I'm seeing the same thing you are seeing with respect to the dot
>>>> sizes. I typically use a value of .0015 or .0020, but changed it to
>>>> .0002 and I don't believe it got any smaller. I was able to easily
>>>> change the density though with cnFillScaleF. (~0.5 gets them
>>>> quite close to one another for a panel plot of mine.)
>>>>
>>>> A NCL developer will have to chime in on where the minimum value of
>>>> cnFillDotSizeF is (assuming we are correct), and if there is a way
>>>> around it.. Best regards, Adam
>>>>
>>>> On 09/30/2012 10:56 PM, mark collier wrote:
>>>>> Hi,
>>>>> I've managed to make the dots smaller (pattern 17) using the
>>>>> cnFillDotSizeF resource but only to a limit. I can change the density
>>>>> of the dots with cnFillScaleF. However, it is still not optimum, I
>>>>> would like to create smaller dots and have them closer together - is
>>>>> this possible? There seems to be a limit do the dot size reduction,
>>>>> although less so with the density.
>>>>>
>>>>> Can a new pattern be created to achieve this?
>>>>>
>>>>> I've attached a figure where I want to make this improvement. I am
>>>>> using these resources with gsn_csm_contour and gsn_contour_shade.
>>>>>
>>>>> Regards,
>>>>> Mark.
>>>>>
>>>>> <Mail Attachment.jpeg>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> ncl-talk mailing list
>>>>> List instructions, subscriber options, unsubscribe:
>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>
>>>> --
>>>> ______________________________________________________________
>>>> Adam Phillips asphilli@ucar.edu
>>>> NCAR/Climate and Global Dynamics Division (303) 497-1726
>>>> P.O. Box 3000
>>>> Boulder, CO 80307-3000 http://www.cgd.ucar.edu/cas/asphilli
>>>> _______________________________________________
>>>> ncl-talk mailing list
>>>> List instructions, subscriber options, unsubscribe:
>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>> _______________________________________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Oct 15 18:14:25 2012