Path: csiph.com!eternal-september.org!feeder.eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail From: Ethan Merritt Newsgroups: comp.graphics.apps.gnuplot Subject: Re: SET FORMAT default Date: Sat, 12 Oct 2019 00:06:25 -0000 (UTC) Organization: A noiseless patient Spider Lines: 18 Message-ID: References: <7c57ce13-c713-6ef3-3690-18d4a35c7efe@t-online.de> <7be2c3cf-1da7-4e8d-aaf5-672c0acb3ada@googlegroups.com> <8c14ccc8-def2-4ed9-8c41-392fb3f76ae3@googlegroups.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Injection-Date: Sat, 12 Oct 2019 00:06:25 -0000 (UTC) Injection-Info: reader02.eternal-september.org; posting-host="13dbbf4819688b526b181ef42d4e1476"; logging-data="310"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX18stgYDpK9j+BFUsLQlq/vX" User-Agent: Pan/0.139 (Sexual Chocolate; GIT bf56508 git://git.gnome.org/pan2) Cancel-Lock: sha1:Kz8QTGmVyLMEe0PnHLxJi1R8fPY= Xref: csiph.com comp.graphics.apps.gnuplot:4244 On Thu, 10 Oct 2019 23:57:34 +0200, Karl Ratzsch wrote: > Am 10.10.2019 um 23:19 schrieb layzarc@aol.com: >> But does that really help if the value being printed comes from a single precision "float"? I guess what I'm asking does Gnuplot use "double" floating point thoughout the SPLOT contour calculation and its printing to the TABLE? > > gnuplot has been using double precison fp math since some time in > the last century I think. > > If the contour calculation doesnt, that would constitute a bug. Do > you have evidence for that? There is one exception that I know of. If a binary matrix is read in, it is stored as (float) rather than (double). There's no good reason for that, but the binary input code is quite tangled so fixing it would require some disentangling. It would be a good project for someone (hint hint). Ethan