AUTOJT: Automatic Job-Typing

Last updated 18 Nov 97

AUTOJT evaluates differences between job descriptions for group pairs. It compares percent time spent and percent members performing values for both groups and prints the results. Up to 850 job description comparisons may be in a single run.

Program Invocation

The specific syntax for program invocation varies between operating systems. This document diplays the code for the AIX version.


@codap autojt <opt> <ssss> <TFf> <TT>

opt:  A   Print report option 1 (PMP)
      B   Print report option 2 (PTS - tasks)
      C   Print report option 3 (performance emphasis)
      D   Print report option 4 (PTS - duties)
      E   Print report option 5 (cumulative PTS - tasks)
      F   Print report option 6 (cumulative PTS - duties)
      G   Print report option 7 (avg. number of tasks performed)
      L   Print additional E-value summary tables
     ' '  If no options A-G are specified, all seven reports are printed.
ssss:     The study number 
TFf:      Input Task Factor file
TT:       Input Task Title or Module Title file

AUTOJT Control Card


AUTOJT <study> <title>

AUTOJT:   cc 01-06  The literal 'AUTOJT'
study:    cc 08-11  The study number for this run
title:    cc 13-72  The report title for this run.

FACTOR Control Card (optional)

The user may optionally weight the differences between group pairs for each task by specifying a task factor on this card. Each task value will be multiplied by the respective value in this factor. The factor values will be printed on the report.


FACTOR <vector> <type>

FACTOR:   cc 01-06  The literal 'FACTOR'
vector:   cc 08-13  The task factor name
type:     cc 15-17  The task factor type

Specification Cards (optional)

The user may optionally change many of the constants used by AUTOJT. There are three cards in this set, with each one being optional. The only requirement is that the cards must be in sequential order (SPEC1 must precede SPEC2 and SPEC3, SPEC2 must precede SPEC3).

SPEC1 Card (optional)

All fields on this card are optional with the exception of the SPEC1 card identifier literal.


SPEC1 <kstat> <ctyp> <isq> <kcon1>... <kcon5> <kcon7>

SPEC1:    cc 01-05  The literal 'SPEC1'
kstat:    cc 07     Flag for option 7 statistical summary report 
                    1 = do not print).
ctyp(1):  cc 09     Constant type for option 1 (D,W,G, default=G)
ctyp(2):  cc 10     Constant type for option 2 (D,W,G*)
ctyp(3):  cc 11     Constant type for option 3 (D,W,G*)
ctyp(4):  cc 12     Constant type for option 4 (D,W,G*)
isq(1):   cc 14     Squaring flag for option 1  (1 = absolute  value,
                    default = square the difference).
isq(2):   cc 15     Squaring flag for option 2
isq(3):   cc 16     Squaring flag for option 3
isq(4):   cc 17     Squaring flag for option 4
kcon(1):  cc 19-26  K-constant for option 1 (default=1.0)
kcon(2):  cc 28-35  K-constant for option 2 (default=1.0)
kcon(3):  cc 37-44  K-constant for option 3 (default=1.0)
kcon(4):  cc 46-53  K-constant for option 4 (default=1.0)
kcon(5):  cc 55-62  K-constant for options 5 + 6 (default=1.0)
kcon(7):  cc 64-71  K-constant for option 7 (default=1.0)

SPEC2 Card (optional)

All fields on this card are optional with the exception of the SPEC2 card identifier literal.


SPEC2 <xcut1> ... <xcut4> <ccon1> ... <ccon5>

SPEC2:    cc 01-05  The literal 'SPEC2'
xcut1:    cc 07-12  Minimum difference cutoff for option 1 (default=0.0)
xcut2:    cc 14-19  Minimum difference cutoff for option 2 (default=0.0)         
xcut3:    cc 21-26  Minimum difference cutoff for option 3 (default=0.0)
xcut4:    cc 28-33  Minimum difference cutoff for option 4 (default=0.0)
ccon1:    cc 35-42  C-constant for option 1 (default=10000.0)
ccon2:    cc 44-51  C-constant for option 2 (default=39.0625)
ccon3:    cc 53-60  C-constant for option 3 (default=9.9)
ccon4:    cc 62-69  C-constant for option 4 (default=156.25)
ccon5:    cc 71-78  C-constant for option 5 (default=12.5)

NOTE: The minimum difference cutoffs are used to restrict tasks that are to be used in calculating the E-value. Option 1 has an additional cutoff for printing tasks which do not enter into the E-value computation (<pmpcut> on the SPEC3 card).

SPEC3 card

All fields on this card are optional with the exception of the SPEC3 card identifier literal.


SPEC3 <dtg1> <dtg2> <dtg3> <dtg4> <g56> <g7> <pmpcut>

SPEC3:    cc 01-05  The literal 'SPEC3'
dtg1:     cc 07-14  Denominator constant for option 1 (default=1.0)
dtg2:     cc 16-23  Denominator constant for option 2 (default=1.0)
dtg3:     cc 25-32  Denominator constant for option 3 (default=1.0)
dtg4:     cc 34-41  Denominator constant for option 4 (default=1.0)
g56:      cc 43-50  G-constant for options 5 and 6 (default=1.0)
g7:       cc 52-59  G-constant for option 7 (default = 1.0)
pmpcut:   cc 61-65  Minimum difference for printing tasks (option 1 only)
                    Default = 20.00

Group Comparison Cards

These cards specify the groups to be compared. Up to 850 group comparisons may be specified, two groups per card. There must be a PTS and a PMP vector available on the input Task Factor file for each group specified.


<group1> <group2>

group1:   cc 01-06  Group ID for this group
group2:   cc 08-13  Group ID for this group

These cards are terminated by an end-of-file or an '@eof' image.

Examples

@codap autojt - ssss tf100 tt001
AUTOJT ssss AUTOJT run for job cluster task comparisons
ST0241 ST0001
ST0196 ST0001
ST0074 ST0031
@eof

Three job comparisons will be computed. All report types are printed using pre-defined cutoffs.

@codap autojt abs sss tf100 tt001
AUTOJT ssss AUTOJT using a task factor for weighting
FACTOR TF0003/DEC
SPEC2  10.00   5.00          7.50
ST0241 ST0001
ST0196 ST0001
ST0074 ST0031
@eof

The report for percent members performing (option 1) will be printed using tasks with a minimum difference of 10.00. The differences will then be weighted, using factor TF0003/DEC as a multiplier for each task. Option 2 and 4 (percent time spent for tasks and duties, respectively) will be printed in a similar fashion, using 5.00 and 7.50 as their respective minimum difference cutoffs.

Back to document index