The following tables lists all of the command line options for the Magnum Opus Command Line system. Parameters are enclosed in square brackets [ and ].
| settings-file=[file name] | Specifies a file that is immediately opened and read. Each line in the file is interpreted as if it were a command line argument. |
At most one of the following may be specified. The default is rule-discovery.
| compile-value-counts | Generates a list of all items or attribute-values together with their respective coverages. |
| find-itemsets | Finds the k itemsets that optimize the search objective within the other constraints that are specified |
| find-rules | Finds the k rules that optimize the search objective within the other constraints that are specified. |
| help | Displays a list of the command line options. |
Each run must include either identifier-item-file, item-list-file, or names-file and data-file commands.
| data-file=[filename] | Specifies the data file for name file-data file data import. |
| delimiter=[character] | Specifies the character that delimits columns in the input file, default is comma. |
| holdout-file=[filename] | Specifies the name of a file from which holdout data should be imported for holdout evaluation. |
| identifier-column=[integer] | The argument is a positive integer that specifies the column of an identifier-item file that contains the identifier. |
| identifier-item-file=[filename] | Specifies the file for identifier-item file import. |
| item-column=[integer] | The argument is a positive integer that specifies the column of an identifier-item file that contains the item. |
| item-list-file=[filename] | Specifies the file for item list file import. |
| names-file=[filename] | Specifies the names file for name file - data file data import. |
| number-of-cases=[ integer] | Specifies the number of cases in an identifier-item, item-list, or data file. Providing this information can reduce data import time. Providing the wrong value can slow data import. |
| ordered | Specifies that an identifier-item file is ordered by identifier. |
| proportion=[no] | Specifies the proportion of cases to be imported. The argument is a number between 0.0 and 1.0. The default is 1.0. |
| save-name-file=[file name] | Specifies the name of a file to which a names file is written after data import. This names file will specify the cut points used for numeric attributes and the values encountered for discrete attributes. |
| csv-file=[filename] | Specifies the name of a file to which rules are output in comma separated value format after rule search is completed. This format is suitable for input into a spreadsheet or other analysis tool for further analysis. |
| filter-mode=[mode] | Specifies the filter mode, allowed values are none, redundant, unproductive, and insignificant, default is insignificant . The depreciated value trivial which is a synonym for redundant is also provided for backwards compatibility. |
| holdout-significance-level=[no] | The argument is a number between 0.0 and 1.0 that specifies the significance level that should be used for holdout evaluation. The default value is 0.05. |
| maximum-results=[integer] | Specifies the number of itemsets to be found. The argument is a positive integer, default = 100. |
| maximum-size=[integer] | Specifies the maximum number of attribute values or items in the left-hand-side of a rule or in an itemset. The argument is a positive integer, default = 4. |
| minimum-coverage=[no] | Specifies the minimum coverage of a rule as a proportion of the total data. The argument is a number greater than 0.0 and less than or equal to 1.0. The default is a value equivalent to 1 case. |
| minimum-coverage-count=[integer] | Specifies the minimum coverage of a rule as a number of cases. The argument is an integer greater than 0. The default is 1. |
| minimum-leverage=[no] | Specifies the minimum leverage of a rule as a proportion of the total data. The argument is a number between 0.0 and 1.0. The default is 0.0. |
| minimum-leverage-count=[integer] | Specifies the minimum leverage of a rule as a number of cases. The argument is a positive integer. The default is 0. |
| no-progress-reports | Specifies that progress reports should be suppressed. |
| out-of-sample-holdout-evaluation | Specifies that the out-of-sample data should be used for holdout evaluation. |
| search-mode=[mode] | Specifies the search mode. Allowed modes are coverage, leverage, lift, strength, and support. The default is leverage. Search by lift and search by strength both require that a constraint be set on at least one of minimum-support, minimum-coverage, or minimum-leverage. If none of these is specified, a default of minimum-support=0.01 is applied. |
| show-coverage=[yes|no] | Specifies whether coverage should be included when a rule is displayed. The default is yes. |
| show-holdout-statistics=[yes|no] | Specifies whether holdout statistics should be displayed for patterns that pass holdout evaluation. The default is no. |
| show-leverage=[yes|no] | Specifies whether leverage should be included when a rule is displayed. The default is yes. |
| show-significance=[yes|no] | Specifies whether significance should be included when a rule is displayed. The default is yes. Note, significance is only displayed when filter-modes insignificant or unsound are used. |
| significance-level=[no] | The argument is a number between 0.0 and 1.0 that specifies the significance level that should be used with the insignificant filter. The default value is 0.01. |
| test-coverage=[yes|no] | Specifies whether holdout evaluation should test for minimum coverage. The default is no. |
| test-leverage=[yes|no] | Specifies whether holdout evaluation should test for minimum leverage. The default is no. |
| test-improvement=[yes|no] | Specifies whether holdout evaluation should test for improvement. The default is no. |
| concise=[1-3] | Sets the format in which rules are displayed. By default rules are displayed in the detailed format. This option is used to select the concise format. The argument 1, 2 or 3 is used to specify the number of lines over which a rule is displayed in the concise format. |
| lhs-available=[list] | Specifies the values that may appear on the left-hand-side of a rule. If a name file is used, specifying an attribute will include all values for that attribute. If lhs-available is used, all values not explicitly included will be excluded. |
| lhs-exclude=[list] | Specifies the values that may not appear on the left-hand-side of a rule. If a name file is used, specifying an attribute will exclude all values for that attribute. If lhs-exclude is used, all values not explicitly excluded will be included. |
| m=[integer] | Specifies that an m-estimate should be used for estimating strength and lift and sets the value of m. The argument is an integer greater than zero. |
| maximum-associations=[integer] | Depreciated form of maximum-results command. |
| maximum-LHS-attrs=[integer] | Depreciated form of maximum-size command. |
| maximum-rules=[integer] | Depreciated form of maximum-results command. |
| minimum-lift=[no] | Specifies the minimum lift for a rule. The argument is a number greater than or equal to 1.0. The default is 1.0. |
| minimum-strength=[no] | Specifies the minimum strength for a rule. The argument is a number between 0.0 and 1.0. The default is 0.0. |
| minimum-support=[no] | Specifies the minimum support of a rule as a proportion of the total data. The argument is a number between 0.0 and 1.0. The default is 0.0. |
| minimum-support-count=[integer] | Specifies the minimum support of a rule as a number of cases. The argument is a positive integer. The default is 0. |
| rhs-available=[list] | Specifies the values that may appear on the right-hand-side of a rule. If a name file is used, specifying an attribute will include all values for that attribute. If rhs-available is used, all values not explicitly included will be excluded. |
| rhs-exclude=[list] | Specifies the values that may not appear on the right-hand-side of a rule. If a name file is used, specifying an attribute will exclude all values for that attribute. If lhs-exclude is used, all values not explicitly excluded will be included. |
| show-lift=[yes|no] | Specifies whether lift should be included when a rule is displayed. The default is yes. |
| show-RHScoverage=[yes|no] | Specifies whether RHScoverage should be included when a rule is displayed. The default is yes. |
| show-strength=[yes|no] | Specifies whether strength should be included when a rule is displayed. The default is yes. |
| show-support=[yes|no] | Specifies whether support should be included when a rule is displayed. The default is yes. |
| test-lift=[yes|no] |
Specifies whether holdout evaluation should test for minimum lift. The default is no. |
| test-partialness=[yes|no] |
Specifies whether holdout evaluation should test for partialness. The default is no. |
| test-positive-correlation=[yes|no] |
Specifies whether holdout evaluation should test for positive correlation. The default is yes. |
| test-strength=[yes|no] |
Specifies whether holdout evaluation should test for minimum strength. The default is no. |
| test-support=[yes|no] | Specifies whether holdout evaluation should test for minimum support. The default is no. |
| items-available=[list] | Specifies the values that may appear in an itemset. If a name file is used, specifying an attribute will include all values for that attribute. If items-available is used, all values not explicitly included will be excluded. |
| items-excluded=[list] | Specifies the values that may not appear in an itemset. If a name file is used, specifying an attribute will exclude all values for that attribute. If items-excluded is used, all values not explicitly excluded will be included. |
Example: mocl name-file=test.nam data-file=test.data
filter-mode=none
outputs the 100 rules with highest leverage from the data in the specified names
file and data file without any filtering.
Each word of an argument may be abbreviated. If there is ambiguity, the first option in alphabetic order is selected.
Example: mocl n-f=test.nam data=test.data f-m=n
is identical to the previous example.
Arguments may be preceded by an optional hyphen (-) or slash (/).
A [character] argument must be a single character.
Values in a [list] argument must be comma separated with no other characters, such as spaces, between items.
| © G I WEBB & ASSOCIATES 1999-2007 | Last updated January 2007 |
| home | products | download | evaluations | prices | purchase | contact us |