Re: [Help-glpk] Need help on interval planning constraint - Basic Bar and Line Graphs (GNU Plot) in GLPK/MathProg

[Harley Mackenzie]

 I have found that the best and most flexible way of plotting large
datasets is to use the open source statistics package R
(http://www.r-project.org/ or download at http://cran.r-project.org).
The R graphics can cope with huge sets of data and there is a wide
variety of graphics styles and add-ons available to use. There are some
good documentation on the web and excellent books also available.

Regards,

Harley

Noli Sicad wrote: On 12/22/09, Nigel Galloway <address@hidden> wrote: 

  When your model is quick to execute then this would be a convinience. If
your model takes several mins to run and you decide a different format is
required for your graph to be clear then it could be annoying having to run
the model again. Saving your result and pasting them into an Excel
spreadsheet, then using its excellent graphic abilities is a good solution. 

 

 The models that I am trying to develop and run have a lot of indexes
from 3 to 6 sets, or more. Forest Management models have thousands of
variables and constraints. It is multiperiod model from 1 to 100 year
time period. It is even hard to interpret the result in glance without
creating graphs. The constraints are non-declining yield, no death in
the plantation, etc. The results needs to be formatted in various
forms for the graphs and output as csv files. I have to run the model
several times to get a realistic outcome.

I am trying to mimic this software - Spatial Woodstock 
http://www.remsoft.com/forestrySoftware.phpOne features of this software are:

* easy to model - themes - set or indexes, I have developed a model
with 10 indexes using this software

* read dbf file or shapefiles for the area (I have done this in
MathProg already)

* creates interactive graphs for easy interpretation of the run.

* expensive - > $25, 0000 for suite initial license - Not my liking
:-). Company pays for this license.

The woodstock model runs in minutes to 1 hour depending on the size of
the model. I have not yet replicated my woodstock model to
GLPK/MathProg yet.

Graphing in Excel is not option while running this big model. Takes a
lot of time to do the graphs in Excel.

I am trying to configure GUSEK to run GNU Plot. I started tweaking /
creating gnuplot.properties for GUSEK and learning GNU Plot. I asked
Luis for help and sent a copy of the draft gnuplot.properties. I think
GUSEK can run GNU Plot

Noli 

  I think one of the glpk examples demonstrates generaring gnuplot commands as
it output. 

 

 _______________________________________________
Help-glpk mailing address@hidden://lists.gnu.org/mailman/listinfo/help-glpk 




-- 
------------------------------------------------------------------
     Dr. Harley Mackenzie         ABN:   36 348 783 012

     HARD Software                Web: www.hardsoftware.comPO BOX 8004          
        Tel:   +61 3 5222 3435
     Newtown 3220, Australia      Email: address@hidden

I have found that the best and most flexible way of plotting large datasets is to use the open source statistics package R (http://www.r-project.org/ or download at http://cran.r-project.org). The R graphics can cope with huge sets of data and there is a wide variety of graphics styles and add-ons available to use. There are some good documentation on the web and excellent books also available.

Regards,

Harley

Noli Sicad wrote:

On 12/22/09, Nigel Galloway <address@hidden> wrote:
  

When your model is quick to execute then this would be a convinience. If
your model takes several mins to run and you decide a different format is
required for your graph to be clear then it could be annoying having to run
the model again. Saving your result and pasting them into an Excel
spreadsheet, then using its excellent graphic abilities is a good solution.
    

The models that I am trying to develop and run have a lot of indexes
from 3 to 6 sets, or more. Forest Management models have thousands of
variables and constraints. It is multiperiod model from 1 to 100 year
time period. It is even hard to interpret the result in glance without
creating graphs. The constraints are non-declining yield, no death in
the plantation, etc. The results needs to be formatted in various
forms for the graphs and output as csv files. I have to run the model
several times to get a realistic outcome.

I am trying to mimic this software - Spatial Woodstock
 http://www.remsoft.com/forestrySoftware.php

One features of this software are:

* easy to model - themes - set or indexes, I have developed a model
with 10 indexes using this software

* read dbf file or shapefiles for the area (I have done this in
MathProg already)

* creates interactive graphs for easy interpretation of the run.

* expensive - > $25, 0000 for suite initial license - Not my liking
:-). Company pays for this license.

The woodstock model runs in minutes to 1 hour depending on the size of
the model. I have not yet replicated my woodstock model to
GLPK/MathProg yet.

Graphing in Excel is not option while running this big model. Takes a
lot of time to do the graphs in Excel.

I am trying to configure GUSEK to run GNU Plot. I started tweaking /
creating gnuplot.properties for GUSEK and learning GNU Plot. I asked
Luis for help and sent a copy of the draft gnuplot.properties. I think
GUSEK can run GNU Plot

Noli








  

I think one of the glpk examples demonstrates generaring gnuplot commands as
it output.
    

_______________________________________________
Help-glpk mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/help-glpk
  

-- 
------------------------------------------------------------------
     Dr. Harley Mackenzie         ABN:   36 348 783 012

     HARD Software                Web:   www.hardsoftware.com
     PO BOX 8004                  Tel:   +61 3 5222 3435
     Newtown 3220, Australia      Email: address@hidden
------------------------------------------------------------------