Scripts

usage: arffviewer.sh [-h]

starts the ArffViewer

 -h   this help

usage: attributes.sh -f <filename> | -d <dirname> [-o <filename>] [-t] [-h]

Retrieves the number of instances from ARFF files

 -h   this help
 -f   <filename>
      the file to check
 -d   <dirname>
      checks all ARFF files in the given directory
      default: ./scripts/../tmp/
 -o   <filename>
      the name of the file to store the output in
 -t   whether to organize the data for one dataset in a single line,
      otherwise the count is printed for each file

usage: change_property.sh -p <property> -v <newvalue> [-f <filename>] [-d <dir>]
       [-b] [-h]

changes a specific property of all ANT files, if '-f <name>' is given
only the property in this file

 -h   this help
 -p   <property>
      the name of the property to change the value for
 -v   <newvalue>
      the new value of the property
 -f   <filename>
      if specified only this file is modified and not all 
 -d   <dir>
      the directory to look in for ANT files
      default: ./scripts
 -b   performs a backup of the original file, i.e. creating a copy
      the extension '.original'

usage: check_ant.sh [-h]

Performs some checks on the ANT XML files, e.g. whether the targets in
./scripts/../database.xml also appear in the other files.

 -h   this help

usage: client.sh -H <hostname> -P <port> -p <port> -o <output> [-h]
       -J <Java-home> -A <ANT-home>

starts the JobClient - the default parameters are listed below.

 -h   this help
 -p   the port on which the client should listen (optional)
      default: 31416
 -H   the hostname/IP of the server to connect to (optional)
      default: localhost
 -P   the port on which the server listens (optional)
      default: 31415
 -o   the directory where to store the logfiles
      default: ./scripts/../tmp
 -v   classname(s) (comma separated list) for which to switch verbose
      mode on (optional)
      default: 
 -A   the location of ANT (above 'bin')
      default: d:/programs/programming/apache-ant-1.6.0
 -J   the location of Java (above 'bin')
      default: d:/programs/programming/jdk/j2sdk1.4.2

usage: copy_arff.sh -s <src-dir> -d <dest-dir> -m <dir> -r <dir> -D -c [-h]

copies arff files to a destination directory (only the data files!)
it also divides them into Multi-Instance and RELAGGS files.

 -h   this help
 -s   <src-dir>
      the directory where the ARFF files are right now
      default: ./scripts/../tmp
 -d   <dest-dir>
      the directory where the ARFF files are to be stored
      default: ./scripts/../tmp/arff
 -m   <dir>
      the sub-directory for the Multi-Instance files
      default: Multi-Instance
 -r   <dir>
      the sub-directory for the RELAGGS files
      default: RELAGGS
 -D   whether to delete the destination dir
      default: no
 -c   compress the files with gzip

usage: datasets.sh -i <datasets> -v [-h]

traverses all README.DATA and stores the data in the README file in
the 'datasets' directory (based on README.skel)

 -h this help
 -i <datasets>
    the directory with the datasets
    default: 

usage: documentation.sh -o <output-dir> -t <type> -d [-h]

generates text or html documentation from all scripts and help.xml

 -h   this help
 -o   <output-dir>
      the directory where to store the documentation
      (it is automatically created if non-existing)
      default: ./scripts/../doc
 -t   <type>
      the type of output, either 'html' or 'text'
      default: html
 -d   deletes the output dir first

usage: dump.sh -o <output-dir> -H <hostname> -P <port> -u <user>
       -p <password> [-c] [-h]

Performs a database dump of the 

 -h   this help
 -o   <output-dir>
      the directory where to store the dump files
      default: ./scripts/../tmp
 -H   <hostname>
      the host where the databases reside
      default: localhost
 -P   <port>
      the port the server listens on
      default: 3306
 -u   <user>
      the name of the user
      default: peter
 -p   <password>
      the password of the user
      default: peter
 -d   <database>
      a specific database to dump (empty means all!)
      default: 
 -c   create CSV files instead of SQL-statements

usage: evaluate.sh -i <input-dir> -o <output-dir> -n <filename> [-h]
       [-p] [-e] [-r] [-t]

creates CSV-files and LaTeX-Tables (german/US-american) from the generated
ARFF stat files

 -h   this help
 -i   <input-dir>
      the directory where the ARFF files are located
      default: ./scripts/../tmp
 -o   <output-dir>
      the directory where to store the CSV/LaTeX files
      default: ./scripts/../tmp
 -n   <filename>
      the filename of the output (CSV/LaTeX)
      default: experiments
 -p   include parameters in file
 -e   check only for exceptions
 -r   check only for runtime
 -t   check only for treesize

usage: filter.sh -f <filtertype> -l <logfile> -a <additional> [-h]

filters logfiles after certain criteria

 -h   this help
 -f   <filertype>
      Current type(s):
      sql      filters Create/Join/Alter/Drop statements
      host     filters all entries from the given  host
               the hostname has to be provided with '-a'
      grep     does a simple grep for the string provided
               with '-a'
 -l   <logfile>
      the logfile to use
 -a   <additional>
      additional parameters, depending on the filtertype

usage: info_client.sh -o <filename> [-h]

Retrieves information about the client and stores it in a text file

 -h   this help
 -o   the file where to store the info
      default: ./scripts/../tmp/info_client.txt

usage: info_clients.sh -H <host-file> -o <filename> [-h]

Retrieves information from clients
For ssh-agent see: http://mah.everybody.org/docs/ssh
The clients file is a simple text file with a hostname in each line.
E.g.:
   somehost
   anotherone

 -h   this help
 -H   the name of the file containing the hostnames for the clients
      default: ./scripts/../tmp/clients
 -o   <filename>
      the file where to store the output
      default: ./scripts/../tmp/info_clients.txt

usage: install.sh -d <dest-dir> -H <db-host> -P <db-port> -u <db-user>
       -p <db-password> -a [-h]

performs the installation of Proper

 -h   this help
 -d   <dest-dir>
      where to install Proper to
      default: /home/fracpete/proper
 -H   <db-host>
      the Datbase host
      default: localhost
 -P   <db-port>
      the port the DB-Server listens on
      default: 3306
 -u   <db-user>
      the user for accessing the DB
      default: peter
 -p   <db-password>
      the password of the user
      default: peter
 -a   performs automatic install without any interaction

usage: jobs.sh -d <dataset> -o <output-file> -H <host> -P <port> [-h]

greps the jobs from ANT XML files and writes them to a file

 -h to display this help
 -d <dataset>
    the dataset(s) to get the jobs from (comma separated list)
    default: 
 -o <output-file>
    the file to store the jobs in
    default: ./scripts/../tmp/jobs
 -H <hostname>
    the hostname of the jdbc url for the database server
    default: localhost
 -P <port>
    the port of the jdbc url for the database server
    default: 3306

usage: mysql.sh -H <hostname> -P <port> -u <user> -p <password> -b [-h]

changes the hostname, port, user and password for MySQL in the ANT files

 -h   this help
 -H   the hostname where the MySQL server resides
      default: localhost
 -P   the port on which the MySQL server listens
      default: 3306
 -u   the user with which to connect to the DB
      default: peter
 -p   the password of the user
      default: peter
 -b   backup the files

usage: performance.sh -i <input-dir> -o <output-dir> -n <filename> [-h]

Retrieves the runtimes from the log files created by the performance.xml
ANT file.

 -h   this help
 -i   <input-dir>
      the directory where the ARFF files are located
      default: ./scripts/../tmp
 -o   <output-dir>
      the directory where to store the CSV/LaTeX files
      default: ./scripts/../tmp
 -n   <filename>
      the filename of the output (CSV/LaTeX)
      default: performance

usage: proper.sh [-c <classname>] [-p <parameter>] [-m <memory>] [-h]

starts the main GUI if no specific class is specified

 -h   this help
 -c   <classname>
      the class to run instead of the main GUI (additional parameters
      can be specified with '-p')
 -p   <parameters>
      additional parameters for other classes besides the main GUI
 -m   <memory>
      amount of memory for the VM, e.g. 800m

usage: run_jobs.sh -j <job-file> [-h]

Runs the jobs generateed for the distributed experiments on a single
machine.

 -h   this help
 -j   the name of the jobfile to execute
      default: ./scripts/../tmp/jobs

usage: scripts.sh -i <interpreter> -b [-h]

changes the interpreter of the scripts

 -h   this help
 -i   the interpreter to use in the script
      default: \/bin\/bash
      you use: /usr/bin/bash
 -b   backup the files

usage: server.sh -j <filename> -r <filename> -p <port> -o <dir> [-h]
       -J <Java-home> -A <ANT-home>

starts the JobServer

 -h   this help
 -j   the filename of the jobs (optional)
      default: ./scripts/../tmp/jobs
 -r   the file where to store the results (optional)
      default: ./scripts/../tmp/results.31415
 -o   the directory to store received files in
      default: ./scripts/../tmp
 -p   the port on which the server should listen (optional)
      default: 31415
 -v   classname(s) (comma separated list) for which to switch verbose
      mode on (optional)
      default: 
 -A   the location of ANT (above 'bin')
      default: d:/programs/programming/apache-ant-1.6.0
 -J   the location of Java (above 'bin')
      default: d:/programs/programming/jdk/j2sdk1.4.2

usage: start_clients.sh -H <host-file> -S <server> -P <server-port>
       -p <client-port> -o <output-dir> [-h]

Starts the JobClients via 'client.sh' and ssh-agent
For ssh-agent see: http://mah.everybody.org/docs/ssh
The clients file is a simple text file with a hostname in each line.
E.g.:
   somehost
   anotherone
Each entry can have its own parameters, i.e. just list the options 
from the command line after a colon ";" in the clients file. E.g.:
   somehost;-A /some/dir/ant
Note: the options -h, -H are not available!

 -h   this help
 -H   the name of the file containing the hostnames for the clients
      default: ./scripts/../tmp/clients
 -S   the hostname/IP of the server to connect to (optional)
      default: localhost
 -P   the port on which the server listens (optional)
      default: 31415
 -p   the port on which the client should listen (optional)
      default: 31416
 -o   the directory where to store the logfiles
      default: ./scripts/../tmp

usage: start_servers.sh -H <host-file> -j <filename> -f <filename>
       -p <client-port> -o <output-dir> [-h]

starts the JobServers via 'server.sh' and ssh-agent
for ssh-agent see: http://mah.everybody.org/docs/ssh
The servers file is a simple text file with a hostname in each line.
E.g.:
   somehost
   anotherone
Each entry can have its own parameters, i.e. just list the options 
from the command line after a colon ";" in the clients file. E.g.:
   somehost;-A /some/dir/ant
Note: the options -h, -H are not available!

 -h   this help
 -H   the name of the file containing the hostnames for the clients
      default: ./scripts/../tmp/servers
 -j   the filename containing the jobs
      default: ./scripts/jobs
 -r   the file where to store the results to
      default: ./scripts/../tmp/results.txt
 -p   the port on which the client should listen (optional)
      default: 31415
 -o   the directory where to store the logfiles
      default: ./scripts/../tmp

usage: stop_clients.sh -H <host-file> [-h]

stops the JobClients via 'kill' and ssh-agent
for ssh-agent see: http://mah.everybody.org/docs/ssh

 -h   this help
 -H   the name of the file containing the hostnames for the clients
      default: ./scripts/../tmp/clients

usage: stop_servers.sh -H <host-file> [-h]

starts the JobServers via 'kill' and ssh-agent
for ssh-agent see: http://mah.everybody.org/docs/ssh

 -h   this help
 -H   the name of the file containing the hostnames for the servers
      default: ./scripts/../tmp/servers