Difference between revisions of "Comparing Run Sets using compare walltime.pl"
| Line 3: | Line 3: | ||
This tutorial assumes that you have already created two sets of data in one or multiple results tables. If they are in the same table they should have a distinct test set labels. | This tutorial assumes that you have already created two sets of data in one or multiple results tables. If they are in the same table they should have a distinct test set labels. | ||
| − | Run the <tt>compare_walltime.pl</tt> script with the proper options<br> | + | Run the <tt>compare_walltime.pl</tt> script with the proper options as they are described below. Running it with no options will give a list of the allowed options<br> |
<br> | <br> | ||
This program will create many .data and .gp files in the current working directory. You must specify a table from which to get the result data. Optionally, a second table may be chosen for comparison. If there is not a second table chosen then there must be both a test set label and second test set label specified. This will compare two particular test sets labels within the same table. Optionally a particular molecule or cluster may also be chosen. | This program will create many .data and .gp files in the current working directory. You must specify a table from which to get the result data. Optionally, a second table may be chosen for comparison. If there is not a second table chosen then there must be both a test set label and second test set label specified. This will compare two particular test sets labels within the same table. Optionally a particular molecule or cluster may also be chosen. | ||
Latest revision as of 15:02, 14 June 2005
compare_walltime.pl
This tutorial assumes that you have already created two sets of data in one or multiple results tables. If they are in the same table they should have a distinct test set labels.
Run the compare_walltime.pl script with the proper options as they are described below. Running it with no options will give a list of the allowed options
This program will create many .data and .gp files in the current working directory. You must specify a table from which to get the result data. Optionally, a second table may be chosen for comparison. If there is not a second table chosen then there must be both a test set label and second test set label specified. This will compare two particular test sets labels within the same table. Optionally a particular molecule or cluster may also be chosen.
Next, run the created gnu plot .gp files. These will automatically choose the correct data files.
gnuplot ./*.gp
This command will create a lot of .png files in the current directory. They are each named <cluster>.<molecule>.png and maybe be viewed with your prefered png viewer.
Clean up is relatively easy. Make sure there are no png, gnuplot or .data files in your current directory and run:rm ./*png ./*data ./*gp
