Classes

Available Classes
Command line Applications (in alphabetical order):
arffcomparer
bagger
classifier
databases
experimenter
export
foreignkeys
import
jobs
job_adder
job_client
job_server
milk
propagation
relaggs
relations
remilk
standalone
tester
GUI Applications (in alphabetical order):
arffviewer
builder
editor
job-monitor
jobber
logger
main
run
setup
sqlviewer
xsler




 proper.app.ArffComparer
 =======================

 Compares two ARFF files and outputs the difference in attributes and in common
 attributes the different instances.
 A prerequisite is that both files have the number of instances.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -file0 <name>
   the first ARFF file
  -file0_type <type>
   the type of the file, currently 'milk' and 'relaggs' are supported
  -attribute0 name[,name[,...]] (optional)
   comma separated list of attribute names to use instead of all
  -file1 <name>
   the second ARFF file
  -file1_type <type>
   see '-file1_type'
  -attribute1 name[,name[,...]] (optional)
   comma separated list of attribute names to use instead of all
  -denote_equal (optional)
   indicates with '[equal]' that both values are the same
  -denote_missing (optional)
   indicates with '[missing]' that both values are missing
  -output <name> (optional)
   if an output file is specified the result is printed to this file in CSV
   format instead of printing it to standard out.




 proper.app.Bagger
 =================

 Creates a Multi-Instance ARFF-File from a table, which can be used with MILK.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -verbose Writer
   switches on the verbose for writing the instances from the database to the ARFF-file

  -table <name>
   the table Relaggs stores the propositionalized data
  -bag <name> (optional)
   the field to use as bag identifier, default is the index in the flattened
   table
  -field <name>
   the field to use as class
  -exclude_fields <name[,name...]> (optional)
   fields you want to exclude from exporting to the classifier (e.g. if they
   contain some implicit information); separate the filed with comma
  -order [yes|no] (optional)
   whether to order fields (default is 'yes')
  -output <name>
   the name of the ARFF-File (used for output)
  -type <classified|unclassified> [optional]
   whether to output only classified instances or only unclassified ones
  -percentage <double> [optional]
   a value between 0 and 100 that represents the percentage of instances to be
   exported (uniformly distributed, i.e. every class is exported with the same
   percentage. 0 or 100 means that all instances are exported.
  -filter <name> (optional)
   the complete name of a WEKA filter
  -filter_additional <args> (optional)
   additional parameters for the filter. if a parameter starts with a '-' it
   has to be written as a '=' instead, otherwise it is interpreted as
   another parameter.
  -missing_binary_to_zero (optional)
   if nominal attributes were postprocessed with the NominalToBinary-Filter
   then the missing values can be set to '0' with this option.




 proper.app.Classifier
 =====================

 Trains a classifier and classifies with it given instances.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -input <name>
   the name of an ARFF-File to build the classifier
  -unknown <name>
   the name of an ARFF-File with instances of unknown classes
  -output <name>
   the name of the ARFF-File to store the prediction in
  -classifier <complete classname>
   the WEKA-/MILK-Classifier's classname
  -additional <add. parameters> (optional)
   additional parameters for the classifier. if a parameter starts with a '-'
   it has to be written as a '=' instead, otherwise it is interpreted as a
   another parameter.




 proper.app.Databases
 ====================

 Creates a Database on the Database server.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -sql_createdatabase <sql> [optional]
   SQL statement that creates a database, e.g. "CREATE DATABASE %NAME%"
  -NOTE about '-database'
   the database(s) to create on the host (comma separated list)
  -initial_database <name> (optional)
   the initial database to connect to, PostgreSQL e.g. does not allow to
   connect to the host alone.




 proper.app.Experimenter
 =======================

 Runs either the WEKA- or the MILK-Experimenter with Crossvalidation.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -input <name>
   the name of an ARFF-File
  -output <name>
   the name of the ARFF-File to store the statistics in
  -classifier <complete classname>
   the WEKA-/MILK-Classifier's classname
  -additional <add. parameters> (optional)
   additional parameters for the classifier. if a parameter starts with a '-'
   it has to be written as a '=' instead, otherwise it is interpreted as
   another parameter.
  -type [cv|loo] [optional]
   the type of evaluation to perform: 'cv' (Crossvalidation) or 'loo' (Leave-
 One-Out) are possible, default is 'cv'
  -runs <number> [optional]
   how often the experiment shall be performed, default is 10
  -folds <number> [optional]
   the number of folds in case of CV, default is 10




 proper.app.Export
 =================

 Exports a table to an ARFF file that can be used with WEKA.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -verbose Writer
   switches on the verbose for writing the instances from the database to the ARFF-file

  -table <name>
   the table Relaggs stores the propositionalized data
  -field <name>
   the field to use as class
  -exclude_fields <name[,name...]> (optional)
   fields you want to exclude from exporting to the classifier (e.g. if they
   contain some implicit information); separate the filed with comma
  -order [yes|no] (optional)
   whether to order fields (default is 'yes')
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"
  -output <name>
   the name of the ARFF-File (used for output)
  -type <classified|unclassified> [optional]
   whether to output only classified instances or only unclassified ones
  -percentage <double> [optional]
   a value between 0 and 100 that represents the percentage of instances to be
   exported (uniformly distributed, i.e. every class is exported with the same
   percentage. 0 or 100 means that all instances are exported.
  -filter <name> (optional)
   the complete name of a WEKA filter
  -filter_additional <args> (optional)
   additional parameters for the filter. if a parameter starts with a '-' it
   has to be written as a '=' instead, otherwise it is interpreted as
   another parameter.
  -missing_binary_to_zero (optional)
   if nominal attributes were postprocessed with the NominalToBinary-Filter
   then the missing values can be set to '0' with this option.




 proper.app.ForeignKeys
 ======================

 Displays the foreign keys of a given table - unfortunately doesn't the current
 MySQL driver supported this.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -table <name>
   the table to retrieve the foreign keys for




 proper.app.Import
 =================

 Used for importing Prolog Databases (or derivates of it) and CSV files
 into a relational database.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -verbose proper.app.Import:1
   switches on the verbose for parsing the file(s)
  -verbose proper.app.Import:2
   switches on the verbose for parsing and the INSERT statements (lots of information!)

  -rel_database <name, [name, ...]>
   the prolog database to import (no explicit IDs, but relational model)
  -background <name, [name, ...]>
   the prolog background knowledge (explicit IDs)
  -add_pos_class (optional)
   adds a positive class label to background knowledge predicates (they are
   considered being positive examples)
  -add_neg_class (optional)
   automatically adds negative examples for missing keys, i.e. if the
   background knowledge covers only some ids then the others are added with
   the negative class label
  -positive <name, [name, ...]>
   the prolog file with positive examples (explicit IDs, no class label)
  -negative <name, [name, ...]>
   the prolog file with negative examples (explicit IDs, no class label)
  -pos_neg <name, [name, ...]>
   the prolog file with positive and negative examples (explicit IDs, negative examples with leading ':- ')
  -unclassified <name, [name, ...]>
   the prolog file with unclassified examples

  -id_index [first|second|third|last|number]... (optional)
   here you can specify whether a certain argument contains the ID (if you)
   you can specify more than one index (separated by comma), e.g. 'first,last'
   if you want to specify it only for a specific predicate, then put it in
   front of the index, separated by a colon, e.g. 'great:last'
   if you want to override a general index for a specific predicate you can do
   with by a preceding '!', e.g. '!bond:first'
   If there is more than one index in one predicate, each instance is split
   into two (the relation between them is via a 'split_id'). If it is not a
   symmetric relationship, you can define this with the following parameter:
  -asymmetric predicate[,predicate[,...]] (optional)
   defines which predicates that have more than one index hava an asymmetric
   relationship between the indices.
  -shuffle (optional)
   randomizes the order of the predicates (to prevent implicit information from
  -reduce_lists (optional)
   whether to reduce lists to regular arguments if they all have the same size
   default: no
   generated keys)
  -list_positions (optional)
   with this flag the positions of the lists added to the table
  -foreign_keys [<predicate1>=<predicate2>[, ...]] (optional)
   the syntax for 'predicate1' and 'predicate2' is like in '-id_index',
   where the predicate1 is the predicate which has a reference to predicate2.
   E.g. a(1,pos). and b(1,blabla). where the first argument in 'a' is a 
   reference to 'b', could be written like this: 
      a:first=b:first
  -parser (optional)
   can be generic (default), or csv
   the csv-parser has the following optional arguments:
    -no_identifiers
     if the file contains no identifiers
    -separator <separator>
     the separator for the fields, default is ','
     Note: 'tab' is interpreted as '\t'
    -qualifier <qualifier>
     the qualifier for strings, default is '"'

 the following combinations of arguments for input files are allowed:
 1. rel_database [, unclassified]
 2. pos_neg [, unclassified]
 3. background, positive, negative [, unclassified]
 4. background, pos_neg [, unclassified]




 proper.app.Jobs
 ===============

 Extracts the Java calls from ANT files and creates jobs for the distributed
 computing.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -input <filename(s)>
   a comma separated list of files to get the java calls from
  -output <filename>
   the name of file to store the jobs in
  -append (optional)
   whether to append the jobs to the given file
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
  -exchange (optional)
   whether to add commands to retrieve and send files
  -target <name[,name...]> (optional)
   create the job files only for this target (comma separated list)
  -no_date (optional)
   this flag suppresses the output of the date comment in the job file




 proper.remote.JobAdder
 ======================

 A little helper for adding Jobs to an already running JobServer.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -server <hostname/IP>
   the server to retrieve jobs from
  -server_port <int> (optional)
   the port to bind to, default is 31415
  -jobs <filename>
   the file containing the jobs to send to the JobServer




 proper.remote.JobClient
 =======================

 A Client waiting for Jobs sent from the JobServer it gregistered itself to.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -port <int> (optional)
   the port to bind to, default is 31416
  -error_log <filename> (optional)
   the file to write errors to, default is errors.log
  -access_log <filename> (optional)
   the file to write accesses to, default is access.log
  -leave_logs (optional)
   doesn't delete the logs when the server starts, default: false
  -accept_timeout <int> (optional)
   the timeout in msec to wait for a connection, default: 30000
  -server <hostname/IP>
   the server to retrieve jobs from
  -server_port <int> (optional)
   the port to bind to, default is 31415
  -output <directory> (optional)
   the directory where to store the logfiles, default is '.'




 proper.remote.JobServer
 =======================

 A Multi-Threaded Server for distributing Jobs to JobClients.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -port <int> (optional)
   the port to bind to, default is 31415
  -error_log <filename> (optional)
   the file to write errors to, default is errors.log
  -access_log <filename> (optional)
   the file to write accesses to, default is access.log
  -leave_logs (optional)
   doesn't delete the logs when the server starts, default: false
  -accept_timeout <int> (optional)
   the timeout in msec to wait for a connection, default: 30000
  -jobs <filename>
   the file where the jobs are stored
  -results <filename>
   where to store the results from the clients
  -append_results
   this option appends the results to the file, otherwise it is emptied first
  -output <directory>
   where to store files received from clients




 proper.app.MILK
 ===============

 Flattens a complete relational database into a single table.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -table <name>
   the table to use as central table for flattening
  -field <name>
   the class field in the central table
  -result_table <name>
   the table Relaggs stores the propositionalized data
  -exclude_tables <name[,name...]> (optional)
   tables you want to exclude from the flattening process
  -join <type> (optional)
   the type of join to perform, available are 'leftouter', 'rightouter',
   'bestouter' (determines the outer join with the least information loss) or
   'natural'
   default is: leftouter
  -max_depth <int> (optional)
   defines the maximal depth the search for relations does, -1 means unlimited
   default is: 1
  -instead_null col=value[,col=value[,...]] (optional)
   defines the values for columns (type is determined automatically) which is
   used instead of NULL values after a Join.
  -tree <tree-string> (optional)
   instead of discovering the structure of the DB itself a tree structure can
   be defined by the user. It starts with the root node, the children are
   separated by ',' and a substree is surrounded by '[' and ']', e.g.:
   a[b,c,d[e,f[g,h]]




 proper.app.Propagation
 ======================

 Propagates the keys to outlying tables, i.e. it creates a star scheme out of
 tree structure.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -table <name>
   the table to use as central table for flattening
  -field <name>
   the class field in the central table
  -exclude_tables <name[,name...]> (optional)
   tables you want to exclude from the flattening process
  -join <type> (optional)
   the type of join to perform, available are 'leftouter', 'rightouter',
   'bestouter' (determines the outer join with the least information loss) or
   'natural'
   default is: leftouter
  -max_depth <int> (optional)
   defines the maximal depth the search for relations does, -1 means unlimited
  -tree <tree-string> (optional)
   instead of discovering the structure of the DB itself a tree structure can
   be defined by the user. It starts with the root node, the children are
   separated by ',' and a substree is surrounded by '[' and ']', e.g.:
   a[b,c,d[e,f[g,h]]




 proper.app.RELAGGS
 ==================

 Performs propositionalization based on RELAGGS, developed by Mark-A. Krogel
 (krogel@iws.cs.uni-magdeburg.de).

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -table <name>
   the table that contains the target
  -result_table <name>
   the table RELAGGS stores the propositionalized data
  -field <name>
   the field to use as class
  -field_in_groupby <name> (optional)
   uses this field additionally in the GROUP BY statements of RELAGGS.
   if the name is omitted it uses '<table>id' as default name
   (might be necessary if classes get lost during the aggregation process)
  -exclude_tables <name[,name...]> (optional)
   tables you want to exclude from the aggregation through Relaggs
  -join <type> (optional)
   can be 'natural' or 'leftouter' ('leftouter' implies 'any_index')
  -any_index (optional)
   whether to use any index or the primary key for aggregation
  -index <name> (optional)
   specifies the index to use instead of the primary key (implies 'any_index')
  -max_depth <int> (optional)
   defines the maximal depth the search for relations does, -1 means unlimited
   default is: 1
  -instead_null col=value[,col=value[,...]] (optional)
   defines the values for columns (type is determined automatically) which is
   used instead of NULL values after a Join.
  -tree <tree-string> (optional)
   instead of discovering the structure of the DB itself a tree structure can
   be defined by the user. It starts with the root node, the children are
   separated by ',' and a substree is surrounded by '[' and ']', e.g.:
   a[b,c,d[e,f[g,h]]




 proper.app.Relations
 ====================

 Displays the relations in a database, starting from one table.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -table <name>
   the table to discover the relations for
  -exclude_tables <name[,name...]> (optional)
   tables you want to exclude from the discovering
  -max_depth <int> (optional)
   defines the maximal depth the search for relations does, -1 means unlimited
   default is: 1




 proper.gui.experiment.ReMilk
 ============================

 A frontend for Propositionalization with REMILK - a combination of RELAGGS
 and MILK.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -filename <name>
   the name of the ANT file to extract the experiments from




 proper.app.Standalone
 =====================

 Transforms the ArffViewer and all necessary components into a standalone
 version with a new package structure. Makes it easy to integrate the Viewer
 in other applications.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -class <name>
   the class from which we shall create a standalone version
  -input_dir <name>
   the directory just above 'proper/gui/...', normally './src'
  -output_dir <name>
   the directory where to copy the transformed files to
  -package <name>
   the new package for all the components, if no package name is provided the
   package declaration is removed from the files.




 proper.app.Tester
 =================

 Tests a previously generated classifier, which can be a WEKA- or MILK
 classifier.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -help
   to print this

  -classifier <complete classname>
   the WEKA-/MILK-Classifier's classname
  -additional <add. parameters> (optional)
   additional parameters for the classifier. if a parameter starts with a '-'
   it has to be written as a '=' instead, otherwise it is interpreted as a
   another parameter.
  -train <filename>
   the filename of the ARFF-file to train the classifier with
  -test <filename>
   the filename of the ARFF-file to test the trained classifier against





 proper.gui.util.ArffViewer
 ==========================

 A little viewer for ARFF files.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this





 proper.gui.experiment.Builder
 =============================

 The Builder enables one to create Proper-specific ANT files for experiments.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this





 proper.gui.util.Editor
 ======================

 A little tool for handling logfiles.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this





 proper.gui.remote.JobMonitor
 ============================

 A Tool for monitoring a JobServer.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -port <int> (optional)
   the port to bind to, default is 31417
  -server <hostname/IP> (optional)
   the server to retrieve jobs from, default is localhost
  -server_port <int> (optional)
   the port to bind to, default is 31415
  -refresh <int> (optional)
  -shell <command> (optional)
   a command that can be executed on the client-tab via the 'Shell'-Button




 proper.gui.remote.Jobber
 ========================

 Creates Jobs for the distributed computing using proper.app.Jobs for this
 purpose.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -file name[,name[,...]]
   the files that shall be processed
   commands like 'synchronize' are prefixed with '#'
   if none is provided then all are used
  -jobfile <filename> (optional)
   the file where to store the resulting jobs, default is jobs
  -exchange (optional)
   whether to include 'get_file' commands or not, default is false
  -append (optional)
   whether to append the jobs to the file or not, default is false




 proper.gui.util.Logger
 ======================

 A little tool for handling logfiles.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this





 proper.gui.Proper
 =================

 The main class containing the menu for starting all other tools.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -height <int> (optional)
   defines the height of the main window in pixel
  -shell <command> (optional)
   a command that can be executed from the Program-Menu




 proper.gui.util.Relations
 =========================

 Prints the relations between tables for a given database and central table.
 These relations can be edited and used for the Propositionaliztion process
 if the automatic discovery is not desired (e.g. to avoid exponential
 explosion of data).

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -driver <name> (optional)
   the driver necessary to access the database
   default is MySQL (org.gjt.mm.mysql.Driver)
  -url <jdbc:subprotocol://host:port/> (optional)
   the url pointing to the host the database resides on
   default is 'jdbc:mysql://localhost:3306/'
  -database <name>
   the database to use on the host
  -user <name> (optional)
   the user to connect with to the database, default is nobody
  -password <pw> (optional)
   the password for the given user
  -sql_limit <sql> [optional]
   SQL statement that limits the result to 1 row, e.g. "LIMIT 1"

  -table <name>
   the table to discover the relations for
  -exclude_tables <name[,name...]> (optional)
   tables you want to exclude from the discovering
  -max_depth <int> (optional)
   defines the maximal depth the search for relations does, -1 means unlimited
   default is: 1




 proper.gui.experiment.Run
 =========================

 Executes any ANT file that was built for Proper, i.e. the predefined ones
 or experiments that were built with the Builder.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -filename <name[,name[,...]]>
   the name of the ANT file(s) to extract the experiments from




 proper.gui.experiment.Setup
 ===========================

 A frontend for setting up the environment for the experiments.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this

  -filename <name>
   the name of the ANT file to extract the experiments from




 proper.gui.util.SqlViewer
 =========================

 A little viewer for SQL queries.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this





 proper.gui.util.XSLer
 =====================

 A little tool for testing XSL scripts. Can also be used for creating Java-
 Code from the generated output.

 Parameters:
  -verbose class[,class[,..]]
   sets the verbose flag of the specified class(es), 'all' matches them all
  -log <filename>
   redirects the output into the specified file
  -append_log [yes|no] (optional)
   with this the file is not deleted at the beginning
  -font <name-face-size> (optional)
   the font to use in lists etc., default is Monospaced-regular-12
  -help
   to print this







Last updated: Tue Jun 8 10:01:02 NZST 2004