Menu
▾
▴
gnuplot-info
|
From: blooopsi <gr...@we...> - 2007-09-19 08:35:42
|
hello there, i need your help. i have a data set of four coulumns and i'd like to plot them with asymmetric yerrorbars. so what i am doing is plot 'dataset' with yerrorbars which should do it. what is happening is, that the errorbars are drawn, but not at the datapoints but somewhere in the graph without datapoints. am i doing anything wrong???? thank you for your help. blooopsi -- View this message in context: http://www.nabble.com/errorbars-tf4479517.html#a12773003 Sent from the Gnuplot - User mailing list archive at Nabble.com. |
|
From: Praedor A. <pr...@ya...> - 2007-09-19 12:56:47
|
On Wednesday 19 September 2007 04:35:46 blooopsi wrote: > hello there, > i need your help. i have a data set of four coulumns and i'd like to plot > them with asymmetric yerrorbars. so what i am doing is > plot 'dataset' with yerrorbars which should do it. what is happening is, > that the errorbars are drawn, but not at the datapoints but somewhere in > the graph without datapoints. am i doing anything wrong???? > thank you for your help. > blooopsi What is the actual command you are sending to gnuplot? =20 =2D-=20 =E2=80=9CThe whole problem with the world is that fools and fanatics are al= ways so=20 certain of themselves, but wiser people so full of doubts.=E2=80=9D - Bertrand Russell |
|
From: blooopsi <gr...@we...> - 2007-09-19 15:29:55
|
the command is plot `agevsgamm.dat` with yerrorbars the data set contains of four columns of numbers so i think that normally it should plot asymmetric errorbars on the datapoints which it isnt doing.it is plotting errorbars but not on my datapoints. they are plotted below the data points which is an absoulte miracle to me why it would do that. thanks for your help :-) blooopsi -- View this message in context: http://www.nabble.com/errorbars-tf4479517.html#a12779883 Sent from the Gnuplot - User mailing list archive at Nabble.com. |
|
From: Praedor A. <pr...@ya...> - 2007-09-19 15:42:29
|
On Wednesday 19 September 2007 11:29:59 blooopsi wrote: > the command is > > plot `agevsgamm.dat` with yerrorbars > > the data set contains of four columns of numbers so i think that normally > it should plot asymmetric errorbars on the datapoints which it isnt > doing.it is plotting errorbars but not on my datapoints. they are plotted > below the data points which is an absoulte miracle to me why it would do > that. > > thanks for your help :-) > blooopsi What is the format of the dat file? Tab separated data? Comma separated? = =20 Are the x y datapoints plotted properly such that only the two error colum= ns=20 are mis-plotted? =2D-=20 =E2=80=9CThe whole problem with the world is that fools and fanatics are al= ways so=20 certain of themselves, but wiser people so full of doubts.=E2=80=9D - Bertrand Russell |
|
From: Praedor A. <pr...@ya...> - 2007-09-19 16:24:53
|
On Wednesday 19 September 2007 11:29:59 blooopsi wrote: > the command is > > plot `agevsgamm.dat` with yerrorbars > > the data set contains of four columns of numbers so i think that normally > it should plot asymmetric errorbars on the datapoints which it isnt > doing.it is plotting errorbars but not on my datapoints. they are plotted > below the data points which is an absoulte miracle to me why it would do > that. Sadly, I am no gnuplot expert. I did what you likely did and read the=20 instructions ("help errorbars", etc). From the gnuplot help information it= =20 is clear that 4 columns SHOULD work as you expect. I quickly threw togethe= r=20 a datafile that contained 4 columns of "data" with the last 2 intended for= =20 minus and plus y error numbers. Plotting thus, 'plot "./test.dat" using=20 1:2:3:4 with yerrorbars' produces a nonsense plot. The x and y data is=20 plotted properly but the two y error columns are, like you experience,=20 plotted separate from the actual xy data (though the plus and minus y error= =20 bars are correct). Same result if I simply do 'plot "./test.dat" with=20 yerrorbars'. Simply changing the command to 'plot "./test.dat" using 1:2:3= =20 with yerrorbars' DID plot properly with the symmetrical y error bars=20 coincident with the datapoints.=20 Someone with much more expertise/experience is going to have to jump in her= e. =20 It seems that gnuplot is behaving contrary to what its in-program help=20 indicates it should. =20 praedor =2D-=20 =E2=80=9CThe whole problem with the world is that fools and fanatics are al= ways so=20 certain of themselves, but wiser people so full of doubts.=E2=80=9D - Bertrand Russell |
|
From: <HBB...@t-...> - 2007-09-19 19:58:24
|
blooopsi wrote: > hello there, > i need your help. i have a data set of four coulumns and i'd like to plot > them with asymmetric yerrorbars. so what i am doing is > plot 'dataset' with yerrorbars which should do it. what is happening is, > that the errorbars are drawn, but not at the datapoints but somewhere in the > graph without datapoints. am i doing anything wrong???? You didn't read 'help yerrorbars' carefully enough. You appear to have x:y:ydelta_up:ydelta_down data, but that's not what gnuplot expects. Change your data or see "help using" for how to have gnuplot change them for you. |
|
From: blooopsi <gr...@we...> - 2007-09-20 08:20:58
|
Thanks a lot for all your help. I finally read the gnuplot help (which i probably should do more often :-)) and solves the problem. for everyone who is interested.it is: plot 'data.dat' using 1:2:($2+$3):($2-$4) with yerrorbars. in all the other gnuplot help sites this isn't necessary. so, i dont know how they produce their errorbars. so thanks again for all your help. blooopsi -- View this message in context: http://www.nabble.com/errorbars-tf4479517.html#a12792816 Sent from the Gnuplot - User mailing list archive at Nabble.com. |
|
From: Praedor A. <pr...@ya...> - 2007-09-20 12:08:24
|
On Thursday 20 September 2007 04:20:57 blooopsi wrote: > Thanks a lot for all your help. I finally read the gnuplot help (which i > probably should do more often :-)) and solves the problem. for everyone w= ho > is interested.it is: > plot 'data.dat' using 1:2:($2+$3):($2-$4) with yerrorbars. > in all the other gnuplot help sites this isn't necessary. so, i dont know > how they produce their errorbars. It doesn't seem that it should be necessary to go all out with the extra co= de =20 since it automatically "knows" that a 3rd column is +/- y error bars if a= =20 3rd column exists and you use the switch "with yerrorbars", it should be=20 assumed that if there are four columns and you use "with yerrorbars", that= =20 should be enough to tell it "columns 3 and 4 are y error data with 3 as=20 positive error and 4 as negative error", etc. Having to use "($2+$3): ($2-$4) " is then redundant to using "with yerrorbars". What ELSE are the= =20 3rd and 4th columns going to be under the circumstances? Same thing if one= =20 uses the switch "with xerrorbars". =20 praedor =2D-=20 =E2=80=9CThe whole problem with the world is that fools and fanatics are al= ways so=20 certain of themselves, but wiser people so full of doubts.=E2=80=9D - Bertrand Russell |
|
From: blooopsi <gr...@we...> - 2007-09-20 12:20:23
|
its the same when i use with xerrorbars, the errorbars then are not on the bottom but on the left hand side of the plot. i really dont know why it is doing this. Praedor Atrebates wrote: >=20 > On Thursday 20 September 2007 04:20:57 blooopsi wrote: >> Thanks a lot for all your help. I finally read the gnuplot help (which i >> probably should do more often :-)) and solves the problem. for everyone >> who >> is interested.it is: >> plot 'data.dat' using 1:2:($2+$3):($2-$4) with yerrorbars. >> in all the other gnuplot help sites this isn't necessary. so, i dont kno= w >> how they produce their errorbars. >=20 > It doesn't seem that it should be necessary to go all out with the extra > code =20 > since it automatically "knows" that a 3rd column is +/- y error bars if > a =20 > 3rd column exists and you use the switch "with yerrorbars", it should be= =20 > assumed that if there are four columns and you use "with yerrorbars", tha= t=20 > should be enough to tell it "columns 3 and 4 are y error data with 3 as= =20 > positive error and 4 as negative error", etc. Having to use "($2+$3): > ($2-$4) " is then redundant to using "with yerrorbars". What ELSE are th= e=20 > 3rd and 4th columns going to be under the circumstances? Same thing if > one=20 > uses the switch "with xerrorbars". =20 >=20 > praedor >=20 > --=20 > =E2=80=9CThe whole problem with the world is that fools and fanatics are = always so=20 > certain of themselves, but wiser people so full of doubts.=E2=80=9D > - Bertrand Russell >=20 > =20 > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Gnuplot-info mailing list > Gnu...@li... > https://lists.sourceforge.net/lists/listinfo/gnuplot-info >=20 >=20 --=20 View this message in context: http://www.nabble.com/errorbars-tf4479517.htm= l#a12796042 Sent from the Gnuplot - User mailing list archive at Nabble.com. |
|
From: Praedor A. <pr...@ya...> - 2007-09-20 13:18:19
|
On Thursday 20 September 2007 08:19:59 blooopsi wrote: > its the same when i use with xerrorbars, the errorbars then are not on the > bottom but on the left hand side of the plot. i really dont know why it is > doing this. > Well, as it works now as you want it with the ($2+$3) thing, I am just=20 throwing out the general question as to WHY this is necessary given that th= e=20 program automatically assumes column 3 is error (both +/-) when one uses=20 the "with (type of)errorbars"...why shouldn't it automatically assume that = a=20 4th, if it exists and one uses "with (type of)errorbars", is thus: x:y: +err:-err? As I stated in my previous post, it appears unnecessarily=20 redundant to state "with yerrorbars" (for instance) AND explicitly tell=20 gnuplot that columns 3 and 4 plus y error and minus y error... It's an honest question. praedor =2D-=20 =E2=80=9CThe whole problem with the world is that fools and fanatics are al= ways so=20 certain of themselves, but wiser people so full of doubts.=E2=80=9D - Bertrand Russell |
|
From: <HBB...@t-...> - 2007-09-20 18:50:43
|
Praedor Atrebates wrote: > Well, as it works now as you want it with the ($2+$3) thing, I am just > throwing out the general question as to WHY this is necessary Because the documentation says so. Period. |
