Graphic file formatsx
Murat Yildizoglu
myildi at gmail.com
Thu Dec 9 00:43:32 UTC 2021
This is not a problem of the text content.
When you create a points plot in R, it puts the information about all the points in the exported PDF figure, even if many of them correspond to the same coordinates. It does not « flatten » the pdf figure. When you plot millions of points, that results in a lot of information and may necessitate considerable amount of memory from the computer for displaying on the screen or from the printer, to transfer on paper.
When we export to a bitmap format like PNG, the information is « flattened » and all points that fall on the same pixels are coded just the value of these pixels. That makes the figure easier to handle.
When you do a curve plot (like a density plot, for example) this problem does not exist. This is another strategy that I use when possible: can I show the same informed with a curve instead of with dots? If yes, I go for it. It is much more parsimonious.
Using a bitmap format with correct proportions and sufficient density, the first solution gives very satisfactory results even if the figures are bitmaps.
--
Prof. Murat Yildizoglu
> Le 9 déc. 2021 à 00:37, Steve Litt <slitt at troubleshooters.com> a écrit :
>
> Rich Shepard said on Wed, 8 Dec 2021 07:51:56 -0800 (PST)
>
>> On Wed, 8 Dec 2021, Dr Eberhard Lisse wrote:
>
>>> For my big handbook (which takes 110 seconds to compile (70 on the
>>> M1)) I have split this into child documents which compile
>>> individually within 10 to 15 seconds.
>>
>> How about loading a PDF image that's 3,704,503 bytes in size while
>> reading your big handbook?
>
> OK, 3,704,503 bytes got my attention. No wonder it's slow.
>
> There are ways to shrink PDF size:
>
> Web search: linux shrink pdf size
>
> SteveT
>
> Steve Litt
> Spring 2021 featured book: Troubleshooting Techniques of the Successful
> Technologist http://www.troubleshooters.com/techniques
> --
> lyx-users mailing list
> lyx-users at lists.lyx.org
> http://lists.lyx.org/mailman/listinfo/lyx-users
More information about the lyx-users
mailing list