Path: csiph.com!news.mixmin.net!weretis.net!feeder4.news.weretis.net!feeder5.news.weretis.net!news.solani.org!.POSTED!not-for-mail
From: Karl Ratzsch <mail.kfr@gmx.net>
Newsgroups: comp.graphics.apps.gnuplot
Subject: Re: consistency of function .vs. data plotting
Date: Mon, 25 Jun 2018 16:06:02 +0200
Organization: solani.org
Lines: 36
Message-ID: <pgqssb$odo$1@solani.org>
References: <a7122b88-09e6-4e90-a094-09d175c8b66b@googlegroups.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
X-Trace: solani.org 1529935563 25016 127.0.0.43 (25 Jun 2018 14:06:03 GMT)
X-Complaints-To: abuse@news.solani.org
NNTP-Posting-Date: Mon, 25 Jun 2018 14:06:03 +0000 (UTC)
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101 Thunderbird/52.8.0
In-Reply-To: <a7122b88-09e6-4e90-a094-09d175c8b66b@googlegroups.com>
Content-Language: en-GB
Cancel-Lock: sha1:kgYj2otpjE17z6E9NAzIMrANKv0=
X-User-ID: eJwFwQEBACAIA7BK4n/UOIDYP4Kb4Oa16HLq6dmGTgLGLtsj7lHNlMlWz4gEk2iIo9yb8QEHXBCJ
Xref: csiph.com comp.graphics.apps.gnuplot:3974

Am 19.06.2018 um 18:41 schrieb palme@kapott.org:

> But the approach using the pseudo data file plots also points for the complex results outside the [-1,1] range.
> 
> See yourself:
>   print "using pure function"
>   plot asin(x)
>   pause -1 "press enter"
>   print "using pseudo data file"
>   plot '+' using 1:(asin($1)) with lines
>   pause -1 "press enter"
> 
> So I have two questions:
> 
> 1) Where exactly is my misunderstanding of the documentation, regarding the way how data file plotting works in comparison to plotting pure functions?
> 
> 2) How can I avoid these complex results to be plotted with the second approach?


gnuplot does not internally report complex values as _undefined_ in data
plots, but does so in function plots. That's imo just a bug.

Of course a data plot containing non-real values makes no real sense,
and you can't even directly read in complex values from a data file,
because there is no predefined format for them.

   set sample 11
   set table
   plot "+" us 1:(1/$1)  # 1/0 _is_ reported as undefined
   plot "+" us 1:(sqrt($1))  # sqrt of negative values isn't

As a workaround, you could do sth like following

   f(x) = ...
   plot "+" us 1:(imag(f(x) =0 ? f(x) : NaN)