gnuplot development

in gnuplot 4.5-cvs (or the coming version 4.4) you
may use the following script:

reset
set xdata time
set timefmt "%Y-%m"
set format x "%m/%Y"
# dummy plot to collect the data set names (plots nothing)
laststring = ""
datasetnames = ""
set table
plot 'datafile' us 1:(stringcolumn(2) eq laststring \
   ? 0 \
   : (laststring = stringcolumn(2) , \
      datasetnames = datasetnames . " " . laststring , \
      0))
unset table
# and now the real plot
plot for [i in datasetnames] 'datafile' \
   using 1:(stringcolumn(2) eq i ? $4 : 0/0) \
   title i with linespoints


Steve Cohen wrote:
> 
> Given the following dataset
> 
> 2009-04    OFF    67935    248967.45    0
> 2009-05    OFF    35693    131338.02    0
> 2009-06    OFF    118    599.07    0
> 2009-10    OFF    1    0.63    0
> 
> 
> 2009-04    NEW    1627    5566.72    1
> 2009-05    NEW    737    4996.18    1
> 2009-06    NEW    414    1902.48    1
> 2009-07    NEW    79    232.22    1
> 2009-08    NEW    119    248.03    1
> 2009-09    NEW    122    190.87    1
> 2009-10    NEW    97    67.92    1
> 
> 
> 2009-05    VERIFIED    1061    6909.02    10
> 2009-06    VERIFIED    3690    27681.30    10
> 2009-07    VERIFIED    1842    13867.37    10
> 2009-08    VERIFIED    1894    13529.98    10
> 2009-09    VERIFIED    1433    10401.02    10
> 2009-10    VERIFIED    1916    10619.17    10
> 
> 
> 2009-06    CERTIFIED    17    185.33    11
> 2009-07    CERTIFIED    163    1067.47    11
> 2009-08    CERTIFIED    127    1037.75    11
> 2009-09    CERTIFIED    118    1148.42    11
> 2009-10    CERTIFIED    532    2301.32    11
> 
> I'd like to get the series titles from Column 2.  Also, the double line 
> feeds between series were added manually.  It would be nice if I didn't 
> have to do that and could trigger a new series based on the contents of 
> column 2.  Are either of these things possible.
> 
> Since I can't do either of these I am doing instead
> 
> set xdata time
> set timefmt "%Y-%m"
> set format x "%m/%Y"
> plot "datafile" index 0 u 1:4 t "OFF" with linespoints, \
> "" index 1 u 1:4 t "NEW" with linespoints, \
> "" index 2 u 1:4 t "VERIFIED" with linespoints,
> "" index 3 u 1:4 t "CERTIFIED" with linespoints
> 
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
> 
> 

-- 
View this message in context: http://www.nabble.com/Series-labels-from-data-file--tp26102735p26114415.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

Great news! Thank you very much. As I'm sure you know, data such as this 
is practically the "canonical" way a SQL aggregate query rolls off. You 
HAVE TO have the repetition in the data. The alternative would somehow 
be to preprocess it (as "report generators" do) or do multiple queries, 
both of which are suboptimal at best.

So the question is, how to get this on the platform I need. I'm pretty 
sure it won't compile on RedHat since they haven't even gone beyond 4.0 
so the question is, is there documentation anywhere for building it with 
all the needed libraries linked in rather than using shared libs? Has 
anyone done this?


Thomas Sefzick wrote:
> in gnuplot 4.5-cvs (or the coming version 4.4) you
> may use the following script:
>
> reset
> set xdata time
> set timefmt "%Y-%m"
> set format x "%m/%Y"
> # dummy plot to collect the data set names (plots nothing)
> laststring = ""
> datasetnames = ""
> set table
> plot 'datafile' us 1:(stringcolumn(2) eq laststring \
>    ? 0 \
>    : (laststring = stringcolumn(2) , \
>       datasetnames = datasetnames . " " . laststring , \
>       0))
> unset table
> # and now the real plot
> plot for [i in datasetnames] 'datafile' \
>    using 1:(stringcolumn(2) eq i ? $4 : 0/0) \
>    title i with linespoints
>
>
> Steve Cohen wrote:
>   
>> Given the following dataset
>>
>> 2009-04    OFF    67935    248967.45    0
>> 2009-05    OFF    35693    131338.02    0
>> 2009-06    OFF    118    599.07    0
>> 2009-10    OFF    1    0.63    0
>>
>>
>> 2009-04    NEW    1627    5566.72    1
>> 2009-05    NEW    737    4996.18    1
>> 2009-06    NEW    414    1902.48    1
>> 2009-07    NEW    79    232.22    1
>> 2009-08    NEW    119    248.03    1
>> 2009-09    NEW    122    190.87    1
>> 2009-10    NEW    97    67.92    1
>>
>>
>> 2009-05    VERIFIED    1061    6909.02    10
>> 2009-06    VERIFIED    3690    27681.30    10
>> 2009-07    VERIFIED    1842    13867.37    10
>> 2009-08    VERIFIED    1894    13529.98    10
>> 2009-09    VERIFIED    1433    10401.02    10
>> 2009-10    VERIFIED    1916    10619.17    10
>>
>>
>> 2009-06    CERTIFIED    17    185.33    11
>> 2009-07    CERTIFIED    163    1067.47    11
>> 2009-08    CERTIFIED    127    1037.75    11
>> 2009-09    CERTIFIED    118    1148.42    11
>> 2009-10    CERTIFIED    532    2301.32    11
>>
>> I'd like to get the series titles from Column 2.  Also, the double line 
>> feeds between series were added manually.  It would be nice if I didn't 
>> have to do that and could trigger a new series based on the contents of 
>> column 2.  Are either of these things possible.
>>
>> Since I can't do either of these I am doing instead
>>
>> set xdata time
>> set timefmt "%Y-%m"
>> set format x "%m/%Y"
>> plot "datafile" index 0 u 1:4 t "OFF" with linespoints, \
>> "" index 1 u 1:4 t "NEW" with linespoints, \
>> "" index 2 u 1:4 t "VERIFIED" with linespoints,
>> "" index 3 u 1:4 t "CERTIFIED" with linespoints
>>
>> ------------------------------------------------------------------------------
>> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
>> is the only developer event you need to attend this year. Jumpstart your
>> developing skills, take BlackBerry mobile applications to market and stay 
>> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
>> http://p.sf.net/sfu/devconference
>> _______________________________________________
>> Gnuplot-info mailing list
>> Gnuplot-info@...
>> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
>>
>>
>>     
>
>

Well, I was able to build, even on RedHat, but it creates a very small 
set of terminals. Missing are any
ubiquitously displayable filetype terminals such as png, gif, or jpeg. I 
also built on Ubuntu, which at least provides png and jpeg. What governs 
this and what can be done about it?


Steve Cohen wrote:
> Great news! Thank you very much. As I'm sure you know, data such as this 
> is practically the "canonical" way a SQL aggregate query rolls off. You 
> HAVE TO have the repetition in the data. The alternative would somehow 
> be to preprocess it (as "report generators" do) or do multiple queries, 
> both of which are suboptimal at best.
>
> So the question is, how to get this on the platform I need. I'm pretty 
> sure it won't compile on RedHat since they haven't even gone beyond 4.0 
> so the question is, is there documentation anywhere for building it with 
> all the needed libraries linked in rather than using shared libs? Has 
> anyone done this?
>
>
> Thomas Sefzick wrote:
>   
>> in gnuplot 4.5-cvs (or the coming version 4.4) you
>> may use the following script:
>>
>> reset
>> set xdata time
>> set timefmt "%Y-%m"
>> set format x "%m/%Y"
>> # dummy plot to collect the data set names (plots nothing)
>> laststring = ""
>> datasetnames = ""
>> set table
>> plot 'datafile' us 1:(stringcolumn(2) eq laststring \
>>    ? 0 \
>>    : (laststring = stringcolumn(2) , \
>>       datasetnames = datasetnames . " " . laststring , \
>>       0))
>> unset table
>> # and now the real plot
>> plot for [i in datasetnames] 'datafile' \
>>    using 1:(stringcolumn(2) eq i ? $4 : 0/0) \
>>    title i with linespoints
>>
>>
>> Steve Cohen wrote:
>>   
>>     
>>> Given the following dataset
>>>
>>> 2009-04    OFF    67935    248967.45    0
>>> 2009-05    OFF    35693    131338.02    0
>>> 2009-06    OFF    118    599.07    0
>>> 2009-10    OFF    1    0.63    0
>>>
>>>
>>> 2009-04    NEW    1627    5566.72    1
>>> 2009-05    NEW    737    4996.18    1
>>> 2009-06    NEW    414    1902.48    1
>>> 2009-07    NEW    79    232.22    1
>>> 2009-08    NEW    119    248.03    1
>>> 2009-09    NEW    122    190.87    1
>>> 2009-10    NEW    97    67.92    1
>>>
>>>
>>> 2009-05    VERIFIED    1061    6909.02    10
>>> 2009-06    VERIFIED    3690    27681.30    10
>>> 2009-07    VERIFIED    1842    13867.37    10
>>> 2009-08    VERIFIED    1894    13529.98    10
>>> 2009-09    VERIFIED    1433    10401.02    10
>>> 2009-10    VERIFIED    1916    10619.17    10
>>>
>>>
>>> 2009-06    CERTIFIED    17    185.33    11
>>> 2009-07    CERTIFIED    163    1067.47    11
>>> 2009-08    CERTIFIED    127    1037.75    11
>>> 2009-09    CERTIFIED    118    1148.42    11
>>> 2009-10    CERTIFIED    532    2301.32    11
>>>
>>> I'd like to get the series titles from Column 2.  Also, the double line 
>>> feeds between series were added manually.  It would be nice if I didn't 
>>> have to do that and could trigger a new series based on the contents of 
>>> column 2.  Are either of these things possible.
>>>
>>> Since I can't do either of these I am doing instead
>>>
>>> set xdata time
>>> set timefmt "%Y-%m"
>>> set format x "%m/%Y"
>>> plot "datafile" index 0 u 1:4 t "OFF" with linespoints, \
>>> "" index 1 u 1:4 t "NEW" with linespoints, \
>>> "" index 2 u 1:4 t "VERIFIED" with linespoints,
>>> "" index 3 u 1:4 t "CERTIFIED" with linespoints
>>>
>>> ------------------------------------------------------------------------------
>>> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
>>> is the only developer event you need to attend this year. Jumpstart your
>>> developing skills, take BlackBerry mobile applications to market and stay 
>>> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
>>> http://p.sf.net/sfu/devconference
>>> _______________________________________________
>>> Gnuplot-info mailing list
>>> Gnuplot-info@...
>>> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
>>>
>>>
>>>     
>>>       
>>   
>>     
>
>
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
>
>
>

you should read about the 'using' command in the manual
http://www.gnuplot.info/docs/node133.html
especially the section about the difference between
'using 1:2' and 'using ($1):($2)'


febty wrote:
> 
> sorry for my late response,
> 
> below is my script :
> 
> #############
> set term postscript enhanced color
> set output "output.ps"
> set lmargin 10
> set nokey
> set ylabel "volt"
> set ytics out
> set xtics ("00:00" 0, "04:00" 18000, "08:00" 32400, "12:00" 46800, "16:00"
> 61200, "20:00" 75600, "23:00" 86400)
> 
> set multiplot
> set size 1, 0.28
> set origin 0.0,0.73
> #set yrange [0.55:0.65]
> #set ytics 0.55, 0.05, 0.65
> set noxlabel
> set title "data"
> set label "channel 1" at screen 0.80, screen 0.94
> set datafile missing "0"
> plot "input.dat" using 0:1 with lines notitle "input.dat"
> 
> set size 1,0.23
> set origin 0.0,0.49
> #set yrange [-0.3:0.1]
> #set ytics -0.3, 0.2, 0.1
> set noxlabel
> set notitle
> set nolabel
> set label "channel 2" at screen 0.80, screen 0.70
> plot "input.dat" using 0:2 with lines notitle "input.dat"
> 
> set size 1,0.23
> set origin 0.0,0.26
> #set yrange [-0.05:0.05]
> #set ytics -0.05, 0.05, 0.05
> set noxlabel
> set notitle
> set nolabel
> set label "channel 3" at screen 0.80, screen 0.47
> plot "input.dat" using 0:3 with lines notitle "input.dat"
> 
> set size 1,0.25
> set origin 0.0,0.0
> #set yrange [-0.1:0.1]
> #set ytics -0.1, 0.1, 0.1
> set xlabel "time (UT)"
> set notitle
> set nolabel
> set label "channel 4" at screen 0.80, screen 0.23
> plot "input.dat" using 0:4 with lines  notitle "input.dat"
> ###########
> 
> I just want to ask, how to plot data without the plot to be interrupted or
> there is a straight  line representing the empty data, if my data has some
> empty data for two line or more.
> 
> 
> best regards,
> febty
> 
> unset multiplot
> 
> 2009/10/24 Thomas Sefzick <t.sefzick@...>
> 
>>
>> maybe you should post the commands you used to plot
>> the data.
>> normally, two empty lines in between cause the plot to
>> be interrupted.
>>
>>
>> febty wrote:
>> >
>> > Dear all,
>> >
>> > If I have data like below
>> >
>> > 1    10
>> > 3   2
>> > 5   3
>> >
>> >
>> > 7   4
>> > 9   5
>> >
>> > How to plot graph so that there is no straight line which connecting
>> > between the four and the fifth column?
>> > I used command set datafile missing "", but there was straight line
>> which
>> > represent the no data column.
>> >
>> > any help are appreciated.
>> >
>> > best regards,
>> > febty
>> >
>>
>> --
>> View this message in context:
>> http://www.nabble.com/plotting-no-data-on-graph-tp26019816p26037301.html
>> Sent from the Gnuplot - User mailing list archive at Nabble.com.
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
>> is the only developer event you need to attend this year. Jumpstart your
>> developing skills, take BlackBerry mobile applications to market and stay
>> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
>> http://p.sf.net/sfu/devconference
>> _______________________________________________
>> Gnuplot-info mailing list
>> Gnuplot-info@...
>> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
>>
> 
> 
> 
> -- 
> 
> ******
> febty febriani
> Indonesian Institute of Sciences
> Research Center for Physics
> Kompleks PUSPIPTEK
> Serpong, Indonesia, 15314
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
> 
> 

-- 
View this message in context: http://www.nabble.com/plotting-no-data-on-graph-tp26019816p26115210.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

'addblanks.awk' assumes that the 1st column contains the
row-values - i.e. fixed value while the values of the 2nd
column are stepped through.
this isn't the case in your data file (2nd column contains
the 'fixed' value).

you need to change '$1' to '$2' in the script:

$2 != prev {printf "\n"; prev=$2} # print blank line


whining_dwarf wrote:
> 
> Hi, I want to generate a pm3d plot the following way:
> 
> set pm3d map;
> splot '<awk -f ~/bin/addblanks.awk  fine_grid_wfs_zint.txt' using 1:2:3;
> 
> using the following file:
> 
>  http://www.nabble.com/file/p26077624/fine_grid_wfs_zint.txt
> fine_grid_wfs_zint.txt 
> 
> The file is sorted and blanks are added as explained in the FAQ. However,
> it does not plot anything.
> Upon unsetting pm3d, I obtain plots as they should be.
> 
> Any hint is appreciated, thanks.
> 

-- 
View this message in context: http://www.nabble.com/pm3d-does-not-draw-anything-tp26077624p26091592.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

you may neglect the date information by dividing the 'seconds
since beginning of the epoch' modulo 24*60*60

set datafile separator ","
set timefmt "%Y-%m-%d %H:%M:%S"
set format x "%H:%M"
plot "datafilename" using (((int(timecolumn(1)) % 86400)/3600)*3600):3


Steve Cohen wrote:
> 
> Suppose I have data structured like this:
> 
> 2009-10-01 10:30:14,16,10,0,0.91,0.29,1.82,1.17
> 2009-10-01 12:30:15,27,17,1,2.13,0.50,3.52,2.58
> 2009-10-01 13:30:17,18,13,0,0.76,0.38,2.07,0.88
> 2009-10-01 14:30:15,22,14,1,1.03,0.32,2.09,1.26
> 2009-10-01 15:30:16,17,11,1,0.95,0.42,2.26,1.11
> 2009-10-01 16:30:16,12,9,0,0.74,0.36,2.05,0.83
> 2009-10-01 17:30:14,14,11,0,0.69,0.24,1.65,0.88
> 2009-10-01 18:30:17,18,15,0,0.80,0.40,2.08,0.93
> 2009-10-01 19:30:15,16,12,0,0.76,0.25,1.69,0.94
> 2009-10-01 20:30:16,14,5,0,0.70,0.19,1.62,0.79
> 2009-10-01 21:30:15,10,7,0,0.44,0.17,1.25,0.55
> 2009-10-01 22:30:15,7,6,0,0.41,0.14,1.30,0.44
> 2009-10-01 23:30:14,4,3,0,0.15,0.09,0.76,0.16
> 2009-10-02 00:30:20,3,2,0,0.15,0.22,0.89,0.08
> 2009-10-02 01:30:14,1,1,0,0.07,0.06,0.53,0.08
> 2009-10-02 02:30:16,2,1,0,3.81,0.98,11.47,0.15
> 2009-10-02 03:30:14,1,1,0,0.07,0.07,0.57,0.03
> 2009-10-02 04:30:14,0,0,0,0.03,0.06,0.35,0.02
> 2009-10-02 05:30:13,1,1,0,0.13,0.08,0.62,0.14
> 2009-10-02 06:30:12,1,0,0,0.13,0.08,0.57,0.14
> 2009-10-02 07:30:13,2,2,0,0.23,0.09,0.81,0.26
> 
> What I want is a plot where the X-axis represents the TIME OF DAY only, 
> without reference to the date, and the Y axes represent various 
> measurements I want to
> at that time over a series of days.  How do I tell Gnuplot to IGNORE the 
> Date portion of the time and only pay attention to the time portion?  In 
> fact, I would not mind losing the minutes and seconds either.  Thus 
> there would be a series of points representing ta y-value at at 1 AM, 2 
> AM, 3 AM, etc. 
> 
> Thanks.
> 
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
> 
> 

-- 
View this message in context: http://www.nabble.com/Plot-time-related-data-by-hour%2C-not-by-day-tp26085157p26085896.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

After studying this some more, I'm thinking my best approach is to 
format my data differently, different queries, etc.

Steve Cohen wrote:
> Suppose I have data structured like this:
>
> 2009-10-01 10:30:14,16,10,0,0.91,0.29,1.82,1.17
> 2009-10-01 12:30:15,27,17,1,2.13,0.50,3.52,2.58
> 2009-10-01 13:30:17,18,13,0,0.76,0.38,2.07,0.88
> 2009-10-01 14:30:15,22,14,1,1.03,0.32,2.09,1.26
> 2009-10-01 15:30:16,17,11,1,0.95,0.42,2.26,1.11
> 2009-10-01 16:30:16,12,9,0,0.74,0.36,2.05,0.83
> 2009-10-01 17:30:14,14,11,0,0.69,0.24,1.65,0.88
> 2009-10-01 18:30:17,18,15,0,0.80,0.40,2.08,0.93
> 2009-10-01 19:30:15,16,12,0,0.76,0.25,1.69,0.94
> 2009-10-01 20:30:16,14,5,0,0.70,0.19,1.62,0.79
> 2009-10-01 21:30:15,10,7,0,0.44,0.17,1.25,0.55
> 2009-10-01 22:30:15,7,6,0,0.41,0.14,1.30,0.44
> 2009-10-01 23:30:14,4,3,0,0.15,0.09,0.76,0.16
> 2009-10-02 00:30:20,3,2,0,0.15,0.22,0.89,0.08
> 2009-10-02 01:30:14,1,1,0,0.07,0.06,0.53,0.08
> 2009-10-02 02:30:16,2,1,0,3.81,0.98,11.47,0.15
> 2009-10-02 03:30:14,1,1,0,0.07,0.07,0.57,0.03
> 2009-10-02 04:30:14,0,0,0,0.03,0.06,0.35,0.02
> 2009-10-02 05:30:13,1,1,0,0.13,0.08,0.62,0.14
> 2009-10-02 06:30:12,1,0,0,0.13,0.08,0.57,0.14
> 2009-10-02 07:30:13,2,2,0,0.23,0.09,0.81,0.26
>
> What I want is a plot where the X-axis represents the TIME OF DAY only, 
> without reference to the date, and the Y axes represent various 
> measurements I want to
> at that time over a series of days.  How do I tell Gnuplot to IGNORE the 
> Date portion of the time and only pay attention to the time portion?  In 
> fact, I would not mind losing the minutes and seconds either.  Thus 
> there would be a series of points representing ta y-value at at 1 AM, 2 
> AM, 3 AM, etc. 
>
> Thanks.
>
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
>
>
>

you may restrict yourself to only one separator character
- actually you have two: "," and whitespace

with only one separator character the plot would be
pretty easy, you would skip the whole date information
and only use time (e.g. whitespace as separator
everywhere):

reset
set xdata time
set timefmt "%H:%M:%S"
set format x "%H:%M"
plot 'datafilename' us 2:3

your data file would look like this:
2009-10-01 10:30:14 16 10 0 0.91 0.29 1.82 1.17

you may even shift your data to full hours if you
like to:
set timefmt "%H"


Steve Cohen wrote:
> 
> After studying this some more, I'm thinking my best approach is to 
> format my data differently, different queries, etc.
> 
> Steve Cohen wrote:
>> Suppose I have data structured like this:
>>
>> 2009-10-01 10:30:14,16,10,0,0.91,0.29,1.82,1.17
>> 2009-10-01 12:30:15,27,17,1,2.13,0.50,3.52,2.58
>> 2009-10-01 13:30:17,18,13,0,0.76,0.38,2.07,0.88
>> 2009-10-01 14:30:15,22,14,1,1.03,0.32,2.09,1.26
>> 2009-10-01 15:30:16,17,11,1,0.95,0.42,2.26,1.11
>> 2009-10-01 16:30:16,12,9,0,0.74,0.36,2.05,0.83
>> 2009-10-01 17:30:14,14,11,0,0.69,0.24,1.65,0.88
>> 2009-10-01 18:30:17,18,15,0,0.80,0.40,2.08,0.93
>> 2009-10-01 19:30:15,16,12,0,0.76,0.25,1.69,0.94
>> 2009-10-01 20:30:16,14,5,0,0.70,0.19,1.62,0.79
>> 2009-10-01 21:30:15,10,7,0,0.44,0.17,1.25,0.55
>> 2009-10-01 22:30:15,7,6,0,0.41,0.14,1.30,0.44
>> 2009-10-01 23:30:14,4,3,0,0.15,0.09,0.76,0.16
>> 2009-10-02 00:30:20,3,2,0,0.15,0.22,0.89,0.08
>> 2009-10-02 01:30:14,1,1,0,0.07,0.06,0.53,0.08
>> 2009-10-02 02:30:16,2,1,0,3.81,0.98,11.47,0.15
>> 2009-10-02 03:30:14,1,1,0,0.07,0.07,0.57,0.03
>> 2009-10-02 04:30:14,0,0,0,0.03,0.06,0.35,0.02
>> 2009-10-02 05:30:13,1,1,0,0.13,0.08,0.62,0.14
>> 2009-10-02 06:30:12,1,0,0,0.13,0.08,0.57,0.14
>> 2009-10-02 07:30:13,2,2,0,0.23,0.09,0.81,0.26
>>
>> What I want is a plot where the X-axis represents the TIME OF DAY only, 
>> without reference to the date, and the Y axes represent various 
>> measurements I want to
>> at that time over a series of days.  How do I tell Gnuplot to IGNORE the 
>> Date portion of the time and only pay attention to the time portion?  In 
>> fact, I would not mind losing the minutes and seconds either.  Thus 
>> there would be a series of points representing ta y-value at at 1 AM, 2 
>> AM, 3 AM, etc. 
>>
>> Thanks.
>>
>> ------------------------------------------------------------------------------
>> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
>> is the only developer event you need to attend this year. Jumpstart your
>> developing skills, take BlackBerry mobile applications to market and stay 
>> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
>> http://p.sf.net/sfu/devconference
>> _______________________________________________
>> Gnuplot-info mailing list
>> Gnuplot-info@...
>> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
>>
>>
>>   
> 
> 
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
> 
> 

-- 
View this message in context: http://www.nabble.com/Plot-time-related-data-by-hour%2C-not-by-day-tp26085157p26090768.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

> attempt to create a gummyworm plot.  You can see the fruits of our labor
> in the attachment.  However, it seems that 'set view map' doesn't take all

Sorry.  I didn't know this list didn't like attachments.

http://mapservice.bco-dmo.org/mapsdev/icons/foo.png

with 'set view map' your plot isn't a real 2d plot, it's
a somehow optimized projection of the 3d plot.
so the margin calculations of the 2d plot do not really
apply here.
'set margin' has effects - e.g. you may set all margins
to zero (without clipping the tic labels - in contrast to
the real 2d plot) to enlarge the plot a little bit, but you
can not reserve space for the title with 'set margin'.

but you may shift your title a little bit in y direction:
set title 'title text' offset 0,3


Charlton Galvarino wrote:
> 
> Hi, all.  Thanks to those who helped me w/ my 1st round of problems in my
> attempt to create a gummyworm plot.  You can see the fruits of our labor
> in the attachment.  However, it seems that 'set view map' doesn't take all
> 3 lines of my title into account, and I end up w/ part of my title
> bleeding over into my graph.  (Had I not 'set view map', the title would
> be a-ok.)
> 
> Setting the margins seems to have no positive effect to control the space
> between the title and the top of the graph.  Is there a margin setting I'm
> missing, or do I need to lose some of my title?
> 
> Many thanks,
> 
> Charlton
> 
> 
> 
> 
> 
> Charlton Galvarino
> 1 (803) 233-6205 : voice
> 1 (803) 223-9579 :   fax
> charlton@...
> 
> P.O. Box 50960
> Columbia, SC 29250
> 
> 
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
> 
> 

-- 
View this message in context: http://www.nabble.com/set-view-map-causes-title-to-bleed-into-graph-tp26078003p26090672.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

you may rename the xtic labels:

set xrange [-10:10]
set xtics ("10" -10, "5" -5, "0" 0, "5" 5, "10" 10)


lyngly wrote:
> 
> Thank you kindly for the helpful reply. This looks promising.
> Your code produces a graph with a range [-10:10], though there will be no
> negative concentrations. Both left and right, there will be a range
> [0:10], I'm wondering how to express that on the graph. Something like
> "xrange [10:0:10]" is not proper syntax I see.
> 
> Best Regards,
> Joseph
> 
> 
> ------------------------------
> 
> Message: 4
> Date: Sat, 24 Oct 2009 11:25:47 -0700 (PDT)
> From: Thomas Sefzick <t.sefzick@...>
> Subject: Re: [Gnuplot-info] Creating Stiff Diagrams
> To: gnuplot-info@...
> Message-ID: <26041711.post@...>
> Content-Type: text/plain; charset=us-ascii
> 
> 
> for a data file like this:
> 
> 2. Na
> 8. Ca
> 9. Mg
> 
> 
> 1. Cl
> 10. HCO3
> 4. SO4
> 
> (i just tried to reproduce the plot on wikipedia.org)
> 
> the plot commands would be:
> 
> reset
> 
> set xrange [-10:10]
> set xtics nomirror
> set yrange [-0.5:2.5]
> set y2range [-0.5:2.5]
> set ytics 1.0 scale 0.0 nomirror
> set y2tics 1.0 scale 0.0 nomirror
> 
> set border 1
> set yzeroaxis linetype -1
> 
> plot "datafilename" using (-$1):0:ytic(2) index 0 with filledcurve x1=0
> linetype 1 notitle, \
> "" using 1:0:y2tic(2) index 1 with filledcurve x1=0 linetype 1 notitle
> 
> or, if the data file would look like this:
> 
> 2. Na 1. Cl
> 8. Ca 10. HCO3
> 9. Mg 4. SO4
> 
> the above plot command could be:
> 
> plot "datafilename" using (-$1):0:ytic(2) with filledcurve x1=0 linetype 1
> notitle, \
> "" using 3:0:y2tic(4) with filledcurve x1=0 linetype 1 notitle
> 
> 
> lyngly wrote:
>> 
>> Hello,
>> 
>> I am new to gnuplot, and am hoping I might be able to use it to create
>> Stiff diagrams (link to wikipedia)--basically I want to plot some
>> irregular hexagons that convey some information about ion composition in
>> waters. My explanation will hopefully make more sense if you first view
>> the example diagram on wikipedia.
>> 
>> Plotting polygons looks fairly simple and straightforward, but only if I
>> arbitrarily insert numbers I need. I am wondering if I can make a custom
>> plot where the x axis is symmetrically bound (0, 20), left and right of
>> 0,
>> e.g. (+20----+10-----0----+10----+20), and the y axis has fixed values
>> for
>> specific ions.  Is there a way to map a text label for the ions to a
>> specific numerical value, and have no intervening numerical values?
>> For instance, if I will always have variables, a, b, c to the left of 0,
>> a*,b*, c* to the right of 0, mapped to the same y-value constants (the
>> ion
>> of interest):
>> (a, -1), (a*, -1*)
>> (b, -8), (b*, -8*)
>> (c, -15), (c*, -15*)
>> 
>> where,
>> -1, -8, -15 are just arbitrary positions for plotting the ion names (3 to
>> left of 0, 3* to the right of 0). I looked through demo graphs and
>> starting browsing through tutorials, but I didn't see anything
>> immediately
>> applicable. Where should I start? Any pointers?
>> 
>> Thank you for reading!
>> 
>> Joseph
>> 
>> 
> 
> 
>       
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry(R) Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart your
> developing skills, take BlackBerry mobile applications to market and stay 
> ahead of the curve. Join us from November 9 - 12, 2009. Register now!
> http://p.sf.net/sfu/devconference
> _______________________________________________
> Gnuplot-info mailing list
> Gnuplot-info@...
> https://lists.sourceforge.net/lists/listinfo/gnuplot-info
> 
> 

-- 
View this message in context: http://www.nabble.com/Re%3A-Creating-Stiff-Diagrams-%28Thomas-Sefzick%29-tp26070657p26073458.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

sorry to clutter everyone's inbox, but here's the .txt file, which 
contains the first set of my data
 uploaded http://www.nabble.com/file/p26061524/pH1.txt pH1.txt 

The columns for the error bars are the sixth and seventh column. 

In addition, I am unable to access the gnuplot help, because my system will
not allow it, so
please forgive me if I'm asking something that is thoroughly covered in the
help section. 
-- 
View this message in context: http://www.nabble.com/Help-with-xyerrorbars-tp26051683p26061524.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

this can not work, because in gnuplot the contents of one record in the
file are separated by whitespace (=one or more blanks or tabs)
this means starting from the 3rd record your data file contains
only 5 values per record:
<pre>
0.046	4.74	0.0901	0.45	0	0.01	0.01
0.086	4.74	0.0762	0	0.225	0.05	0.001
0.129	4.74	0.0586			0.002	0.001
0.146	4.74	0.0511			0.05	0.001
</pre>

set datafile missing "?"
and produce a new data file with this missing character
at the place of a missing values:
<pre>
0.046	4.74	0.0901	0.45	0	0.01	0.01
0.086	4.74	0.0762	0	0.225	0.05	0.001
0.129	4.74	0.0586	?	?	0.002	0.001
0.146	4.74	0.0511	?	?	0.05	0.001
</pre>

or,
set datafile separator "\t"
this should work because the values in your file are separated by
tab characters.


Noora wrote:
> 
> sorry to clutter everyone's inbox, but here's the .txt file, which 
> contains the first set of my data
>  uploaded http://www.nabble.com/file/p26061524/pH1.txt pH1.txt 
> 
> The columns for the error bars are the sixth and seventh column. 
> 
> In addition, I am unable to access the gnuplot help, because my system
> will not allow it, so
> please forgive me if I'm asking something that is thoroughly covered in
> the help section. 
> 

-- 
View this message in context: http://www.nabble.com/Help-with-xyerrorbars-tp26051683p26062409.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.