Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.graphics.apps.gnuplot > #881 > unrolled thread
| Started by | Gudrun <bestenborstel@gmail.com> |
|---|---|
| First post | 2012-01-31 17:31 +0000 |
| Last post | 2012-02-09 05:43 +0100 |
| Articles | 6 — 3 participants |
Back to article view | Back to comp.graphics.apps.gnuplot
Plotting a range accorrding to given numbers Gudrun <bestenborstel@gmail.com> - 2012-01-31 17:31 +0000
Re: Plotting a range accorrding to given numbers Hans-Bernhard Bröker <HBBroeker@t-online.de> - 2012-01-31 22:31 +0100
Re: Plotting a range accorrding to given numbers Gudrun <bestenborstel@gmail.com> - 2012-02-03 15:59 +0000
Re: Plotting a range accorrding to given numbers Hans-Bernhard Bröker <HBBroeker@t-online.de> - 2012-02-04 18:43 +0100
Re: Plotting a range accorrding to given numbers Gudrun <bestenborstel@gmail.com> - 2012-02-06 12:07 +0000
Re: Plotting a range accorrding to given numbers Hermann Peifer <peifer@gmx.eu> - 2012-02-09 05:43 +0100
| From | Gudrun <bestenborstel@gmail.com> |
|---|---|
| Date | 2012-01-31 17:31 +0000 |
| Subject | Plotting a range accorrding to given numbers |
| Message-ID | <jg98if$kbi$1@news.albasani.net> |
Dear all Let's assume you have data given in such a form x y yerr 1 2 12 1 2 10 10 1 3 20 13 2 4 30 14 1 The first column is basically the counter/index for all other columns. Can I tell gnuplot somehow to plot x(1:3), meaning plotting points 1 to 3 for columns x and also for y and yerr? I don't want to take care about the values, I would like to assign this by index. Reason: Sometimes it is a bit hard to plot 2<x<20 if the numbers are very close together and have a lot of decimal places behind the point. Thank you for your help. Kind regards Gudrun
[toc] | [next] | [standalone]
| From | Hans-Bernhard Bröker <HBBroeker@t-online.de> |
|---|---|
| Date | 2012-01-31 22:31 +0100 |
| Message-ID | <9or4tsFmgjU2@mid.dfncis.de> |
| In reply to | #881 |
On 31.01.2012 18:31, Gudrun wrote: > Dear all > > Let's assume you have data given in such a form > x y yerr > 1 2 12 1 > 2 10 10 1 > 3 20 13 2 > 4 30 14 1 > > The first column is basically the counter/index for all other columns. > Can I tell gnuplot somehow to plot x(1:3), meaning plotting points 1 to > 3 for columns x and also for y and yerr? I don't want to take care about > the values, I would like to assign this by index. You're not being terribly clear, but it seems like you might be searching for 'help every', particularly plot 'file' every ::0::2
[toc] | [prev] | [next] | [standalone]
| From | Gudrun <bestenborstel@gmail.com> |
|---|---|
| Date | 2012-02-03 15:59 +0000 |
| Message-ID | <jgh08u$7m4$1@news.albasani.net> |
| In reply to | #884 |
Dear Hans-Bernhard, sorry for my mis-explanation. x is the second column in the example below, y the third. The first column is the index. I just want to plot x-column, y-column, and error-column with the help of the index-column. Let's say index_1=2 (second line) and index_2=4 (last line). Is there a command to plot x(index_1..index2) vs. y(index_1..index_2), meaning that the following is only plotted: x y 2 10 3 20 4 30 Is this now by any chance better explained? Sorry. Thank you for your help. G. Am 31.01.12 21:31, schrieb Hans-Bernhard Bröker: > On 31.01.2012 18:31, Gudrun wrote: >> Dear all >> >> Let's assume you have data given in such a form >> x y yerr >> 1 2 12 1 >> 2 10 10 1 >> 3 20 13 2 >> 4 30 14 1 >> >> The first column is basically the counter/index for all other columns. >> Can I tell gnuplot somehow to plot x(1:3), meaning plotting points 1 to >> 3 for columns x and also for y and yerr? I don't want to take care about >> the values, I would like to assign this by index. > > You're not being terribly clear, but it seems like you might be > searching for 'help every', particularly > > plot 'file' every ::0::2 >
[toc] | [prev] | [next] | [standalone]
| From | Hans-Bernhard Bröker <HBBroeker@t-online.de> |
|---|---|
| Date | 2012-02-04 18:43 +0100 |
| Message-ID | <9p591rFh1tU1@mid.dfncis.de> |
| In reply to | #903 |
On 03.02.2012 16:59, Gudrun wrote: > I just want to plot x-column, y-column, and error-column with the help > of the index-column. The nature of that "help" is still a bit unclear, because the index column in your example file is deceptively simple. As-is, it's actually superfluous, because it's just the line number. So the 'every' option I mentioned before could actually do what you want in this particular case. The question remains what is supposed to happen if the index column were, say 0 5 1 3 2 4 or 5 4 3 2 1 0
[toc] | [prev] | [next] | [standalone]
| From | Gudrun <bestenborstel@gmail.com> |
|---|---|
| Date | 2012-02-06 12:07 +0000 |
| Message-ID | <jgofp7$e5l$1@news.albasani.net> |
| In reply to | #905 |
Good morning, Hans-Bernhard. I am confused with "every" and "index". I thought index is the command I am looking for. In the end I am looking for a way to choose by line number if you wish to say so. Every seems to work fine, although Gnuplot makes funny things if the entry is NaN. Kind regards Gudrun Am 04.02.12 17:43, schrieb Hans-Bernhard Bröker: > On 03.02.2012 16:59, Gudrun wrote: > >> I just want to plot x-column, y-column, and error-column with the help >> of the index-column. > > The nature of that "help" is still a bit unclear, because the index > column in your example file is deceptively simple. As-is, it's actually > superfluous, because it's just the line number. So the 'every' option I > mentioned before could actually do what you want in this particular case. > > The question remains what is supposed to happen if the index column > were, say > > 0 > 5 > 1 > 3 > 2 > 4 > > or > > 5 > 4 > 3 > 2 > 1 > 0 >
[toc] | [prev] | [next] | [standalone]
| From | Hermann Peifer <peifer@gmx.eu> |
|---|---|
| Date | 2012-02-09 05:43 +0100 |
| Message-ID | <jgvitc$mhs$1@news.albasani.net> |
| In reply to | #905 |
On 04/02/2012 18:43, Hans-Bernhard Bröker wrote: > On 03.02.2012 16:59, Gudrun wrote: > >> I just want to plot x-column, y-column, and error-column with the help >> of the index-column. > > The nature of that "help" is still a bit unclear, because the index > column in your example file is deceptively simple. As-is, it's actually > superfluous, because it's just the line number. So the 'every' option I > mentioned before could actually do what you want in this particular case. > > The question remains what is supposed to happen if the index column > were, say > > 0 > 5 > 1 > 3 > 2 > 4 > > or > > 5 > 4 > 3 > 2 > 1 > 0 > I would suggest something like: plot "<( awk '$1 > 0 && $1 <= 3' my.dat)" ... The above would ignore the order of the values in the first column and also plot in case of multiple occurrences of 1, 2 or 3. There are other AWK-Oneliners for only plotting the first occurrence of something, the last one, a random one, etc. Occasionally, people in this group recommend to do data pre-processing with AWK and similar tools and use gnuplot for plotting only. I can only agree. Hermann
[toc] | [prev] | [standalone]
Back to top | Article view | comp.graphics.apps.gnuplot
csiph-web