Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.graphics.apps.gnuplot > #4171 > unrolled thread

plot datafile with a blank between +- and some data

Back to article view | Back to comp.graphics.apps.gnuplot

Contents

  plot datafile with a blank between +- and some data Jörg Buchholz <bookwood4new@freenet.de> - 2019-04-03 08:38 +0200
    Re: plot datafile with a blank between +- and some data Jörg Buchholz <bookwood4new@freenet.de> - 2019-04-03 20:33 +0200
      Re: plot datafile with a blank between +- and some data Karl Ratzsch <mail.kfr@gmx.net> - 2019-04-03 22:06 +0200
        Re: plot datafile with a blank between +- and some data Jörg Buchholz <bookwood4new@freenet.de> - 2019-04-04 10:10 +0200
          Re: plot datafile with a blank between +- and some data Karl Ratzsch <mail.kfr@gmx.net> - 2019-04-04 23:47 +0200

#4171 — plot datafile with a blank between +- and some data

Hello,

how can I plot the following data structure out of a file?

33,2019/03/21 10:00:06,  0,+ 6.786,+ 2.507
34,2019/03/21 10:00:07,  0,+10.089,+ 5.564
35,2019/03/21 10:00:08,  0,- 0.584,+ 3.324
36,2019/03/21 10:00:09,  0,+10.156,+ 0.024
37,2019/03/21 10:00:10,  0,+ 4.278,+ 5.439
38,2019/03/21 10:00:11,  0,+ 0.431,+ 0.262


set datafile separator comma
plot 'datafile' u 1:4 w p pt 7

gives only the values higher than 10 to me, cause there are no space
between the sign and the value.

How can I set the format correct to get all points?

thanks

Jörg

[toc] | [next] | [standalone]

#4172

On 03.04.19 08:38, Jörg Buchholz wrote:
> Hello,
> 
> how can I plot the following data structure out of a file?
> 
> 33,2019/03/21 10:00:06,  0,+ 6.786,+ 2.507
> 34,2019/03/21 10:00:07,  0,+10.089,+ 5.564
> 35,2019/03/21 10:00:08,  0,- 0.584,+ 3.324
> 36,2019/03/21 10:00:09,  0,+10.156,+ 0.024
> 37,2019/03/21 10:00:10,  0,+ 4.278,+ 5.439
> 38,2019/03/21 10:00:11,  0,+ 0.431,+ 0.262
> 
> 
> set datafile separator comma
> plot 'datafile' u 1:4 w p pt 7
> 
> gives only the values higher than 10 to me, cause there are no space
> between the sign and the value.
> 
> How can I set the format correct to get all points?

At the moment my solution is the following:

system "sed -i 's/+ / +/g;s/- / -/g' datafile"

Is there a solution to do it inside gnuplot?

Jörg

[toc] | [prev] | [next] | [standalone]

#4174

Am 03.04.2019 um 20:33 schrieb Jörg Buchholz:
> On 03.04.19 08:38, Jörg Buchholz wrote:
>> Hello,
>>
>> how can I plot the following data structure out of a file?
>>
>> 33,2019/03/21 10:00:06,  0,+ 6.786,+ 2.507
>> 34,2019/03/21 10:00:07,  0,+10.089,+ 5.564
>> 35,2019/03/21 10:00:08,  0,- 0.584,+ 3.324
>> 36,2019/03/21 10:00:09,  0,+10.156,+ 0.024
>> 37,2019/03/21 10:00:10,  0,+ 4.278,+ 5.439
>> 38,2019/03/21 10:00:11,  0,+ 0.431,+ 0.262
>>
>>
>> set datafile separator comma
>> plot 'datafile' u 1:4 w p pt 7
>>
>> gives only the values higher than 10 to me, cause there are no space
>> between the sign and the value.
>>
>> How can I set the format correct to get all points?
> 
> At the moment my solution is the following:
> 
> system "sed -i 's/+ / +/g;s/- / -/g' datafile"
> 
> Is there a solution to do it inside gnuplot?

Possibly, but it will be very cumbersome. Putting a space between
the sign and the decimal number is really a mean trick.

"plot" can use a scanf() format specifier, but it cannot handle the
date/time column, and only seven values in total, plus it would
still stumble over the space in column 4.

Be that as it may:

$dat << EOD
33,2019/03/21 10:00:06,  0,+ 6.786,+ 2.507
34,2019/03/21 10:00:07,  0,+10.089,+ 5.564
35,2019/03/21 10:00:08,  0,- 0.584,+ 3.324
36,2019/03/21 10:00:09,  0,+10.156,+ 0.024
37,2019/03/21 10:00:10,  0,+ 4.278,+ 5.439
38,2019/03/21 10:00:11,  0,+ 0.431,+ 0.262
EOD

set dataf sep "X"

col1(st) = st[1:2] * 1.0
col4(st) = (st[28:28] eq "-" ? -1 : 1) * st[29:34]
stcol(n) = stringcolumn(n)

set table
plot $dat us (col1(stcol(1))):(col4(stcol(1))) with table
unset table

[toc] | [prev] | [next] | [standalone]

#4175

On 03.04.2019 22:06, Karl Ratzsch wrote:
> Am 03.04.2019 um 20:33 schrieb Jörg Buchholz:
>> On 03.04.19 08:38, Jörg Buchholz wrote:
>>> Hello,
>>>
>>> how can I plot the following data structure out of a file?
>>>
>>> 33,2019/03/21 10:00:06,  0,+ 6.786,+ 2.507
>>> 34,2019/03/21 10:00:07,  0,+10.089,+ 5.564
>>> 35,2019/03/21 10:00:08,  0,- 0.584,+ 3.324
>>> 36,2019/03/21 10:00:09,  0,+10.156,+ 0.024
>>> 37,2019/03/21 10:00:10,  0,+ 4.278,+ 5.439
>>> 38,2019/03/21 10:00:11,  0,+ 0.431,+ 0.262
>>>
>>>
>>> set datafile separator comma
>>> plot 'datafile' u 1:4 w p pt 7
>>>
>>> gives only the values higher than 10 to me, cause there are no space
>>> between the sign and the value.
>>>
>>> How can I set the format correct to get all points?
>>
>> At the moment my solution is the following:
>>
>> system "sed -i 's/+ / +/g;s/- / -/g' datafile"
>>
>> Is there a solution to do it inside gnuplot?
> 
> Possibly, but it will be very cumbersome. Putting a space between
> the sign and the decimal number is really a mean trick.

It is a .csv file that comes from a "Graphtec GL240 midi Data Logger".

> "plot" can use a scanf() format specifier, but it cannot handle the
> date/time column, and only seven values in total, plus it would
> still stumble over the space in column 4.
> 
> Be that as it may:
> 
> $dat << EOD
> 33,2019/03/21 10:00:06,  0,+ 6.786,+ 2.507
> 34,2019/03/21 10:00:07,  0,+10.089,+ 5.564
> 35,2019/03/21 10:00:08,  0,- 0.584,+ 3.324
> 36,2019/03/21 10:00:09,  0,+10.156,+ 0.024
> 37,2019/03/21 10:00:10,  0,+ 4.278,+ 5.439
> 38,2019/03/21 10:00:11,  0,+ 0.431,+ 0.262
> EOD
> 
> set dataf sep "X"
> 
> col1(st) = st[1:2] * 1.0
> col4(st) = (st[28:28] eq "-" ? -1 : 1) * st[29:34]
> stcol(n) = stringcolumn(n)
> 
> set table
> plot $dat us (col1(stcol(1))):(col4(stcol(1))) with table
> unset table


Thanks for your solution. There are some more columns. So I think, for
me, it is easer do handle the datafile with sed. From inside gnuplot or
direct after the files where produced.

Jörg

[toc] | [prev] | [next] | [standalone]

#4176

Am 04.04.2019 um 10:10 schrieb Jörg Buchholz:
> On 03.04.2019 22:06, Karl Ratzsch wrote:
>> col1(st) = st[1:2] * 1.0
>> col4(st) = (st[28:28] eq "-" ? -1 : 1) * st[29:34]
>> stcol(n) = stringcolumn(n)
>>
>> set table
>> plot $dat us (col1(stcol(1))):(col4(stcol(1))) with table
>> unset table
> 
> 
> Thanks for your solution. There are some more columns. So I think, for
> me, it is easer do handle the datafile with sed.

Me too. If the format is not totally fixed like in your example,
this becomes bedlam with gnuplot string evaluation. Certainly
possible, but it is probably easier to learn regular expressions. :-)

[toc] | [prev] | [standalone]

Back to top | Article view | comp.graphics.apps.gnuplot

csiph-web