DISTRO II-DISTRIBUTION ROTATION MODEL Joanne M. Witt Army Research Institute for the Behavioral and Social Sciences Arlington, Virginia April 1973

Transcription

1 AD DISTRO II-DISTRIBUTION ROTATION MODEL Joanne M. Witt Army Researh Institute for the Behavioral and Soial Sienes Arlington, Virginia April 1973 DISTRIBUTED BY: mi] National Tehnial Information Servie U. S. DEPARTMENT OF OMMERE 5285 Port Royal Road, Springfield Va

2 Ttiiiioi Researh Report 1176 AD DISTRO ll.-distribution ROTATION MODEL» Joanne M Witt O a < Pauline T. Olson, Work Unit LMdW rprilulrd hv NATIONAL TEHNIAL INFORMATION SERVIE U S Deporlmjnt o( nnm.i.tr» '.,,..g«...;,! VA.'.'IM n MAY 3 2 JU JuEE i J U. S. Army Army Researh Institute for the Behavio.dl and Soial Sienes April 1973 Approved tor pub'i release, distribution uniimiied.

3 Unlassified Seurity U»»ifi»tion DOUMENT ONTROL DATA RID (Smutlir laitltlulon ol llll; boiy ol»btltmel mnd Induing mnouiton muh b» «nl«mrf whit Iht onrmll npotl It l»$tlll»4j I. oniinatino ATIVITY fmpsnl* wthor; U. S. Army Researh Institute for the Behavioral and Soial Sienes, Arlington, VA. to. RPOIIT SIURITV LAMIFI * TION Unlassified i*. OMOUP a. ntponr TITLI DISTRO ll-distribution ROTATION MODEL 4. DltRiPTivi NOTi (Typ* ol npotl and Inh»*** dmf) t «uthoniti (Flnlnamo, mimio Inlllol, laolnom») Joanne M. Witt «. mtpomj OATI April 1973 M. ONTRAT OH GRANT NO. fe. RROJT NO. DA R&D Proj No. 2O065101M746 SIMPO A lib 70. TOTAL NO. Or»ASI 61 ONIOINATOR't RRORT NUMVKRtai Tehnial Researh Report 1176 OTHIR RIRORT NOI1I (Any outot HI uatbon Mat may b* a*. ifil» npotl) Approved for publi release: distribution unlimited. II.»URRLtMNTARV NOT!«II ironiorino MILITARY ATIVITY Offie, Deputy hief of Staff for Personnel, Wash., D Offie of Personnel Operation, Wash., D IS. AeiTRAT / In a ontinuation of effort under in operations researh requirement, the Statistial Researh and Analysis Division. BESRL has been engaged mstudy and evaluation of the Army's personnel system through development of a model enabln <) A^mv personnel management to assess quantitatively the impat of poliies on deployment and readiness. The present report deals with the development and use of DISTRO-II (Distribution Rotation model), essentially a modifiation of the General Matrix Manipulator (GMM), a previous BESRL model. DISTRO-II, whih simulates additional fators affeting personnel distribution, more aurately models deployability harateristis of the Army than does either the GMM or DISTRO I The DISTRO-II model is desribed in the management setion of the report. Priniples of the GMM and its mterteoe with DISTRO II are presented in detail In the tehnial supplement setion. FORTRAN listings of DISTR02 subroutines are provided in the appendix. ^n PMM 4 A ^4 R«RLAeti oe rorm»«t». I JAM M ÜÜ i MOV MI4 / O kiti re«ar*»y ua*. MIM ta Unlassified Seurity UatlftaUon

4 Unlassitied Seurity liitlfitlion KEY wonof Model simulation Model simulation *omputer tehnology *Automated systems *Personnel management systems *Model appliation *Modei development Deployability; Nondeployability *DISTRO-ll (Distribution Rotation model) Personnel poliy evaluation *Programming language Node; node flows System requirements Resoures GMM (General Matrix Manipulator) Matries Interfae /I/ Unlassified Seurity Uttlflation

5 Tehniol Researh Report 1176 AD DISTRO II--DISTRIBUTION ROTATION MODEL Joanne M Witt Pauline T. Olson, Woik Unit Lndll STATISTIAL RESEARH AND ANALYSIS DIVISION eil D Johnson, hief ARMY RESEARH INSTITUTE FOR THE BEHAVIORAL AND SOIAL SIENES Offie, hief of Researh and L./elopment Department of the Army 1300 Wilson Boulevard, Ailimiton. Virginia April 1973 \V Projet Number Q065101M746 Appfowd (or puhlir Nttnt; dislnhulion unlimitil / SIMPO II I)

6 ARI Tehnial Researh Reports and Tehnial Researh Notes are intenrieri for sponsors of R&D tasks and other researh and military agenies. Any findings ready for implementation at the time of publiation are presented in the latter part of the Brief. Upon ompletion of a major phase of the task, formal reommendations for offiial ation normally are onveyed to appropriate military agenies by briefing or Disposition Form. 00

7 FOREWORD The researh reported ho.e was aompli^ied by the U. S. Army Researh Institute for the behavioral and Soial Sienes (ARII. The Institute, established 1 Otober 1972 as replm ^tient for the U. S. Army Manpower Resoures Researh and Development enter, unifies in one mlarged organization all ORD ativities in the behavioral and soial siene area, inluding those formerly onduted by the Behavior and Systems Researh Laboratory (BESRU and the Motivation and Training Laboratory (MTL). The present Tehnial Researh Report was prepared while BESRL existed as a separate entity and reflets Division and Work Unit struture as onstituted prior to 1 Otober The Work Unit "Simulation Models of Personnel Operations II SIMPO" was a ontinuation of effort under an operations researh requirement desribed in the Army Master Study Program under the title "A Simulation Model of Personnel Operations (SIMPO ) and isprojet2q065101m746(fy 1972), "Army Operations and Intelligene Analysis "under the auspies of the Army Study Advisory ommittee. The present Tehnial Researh Report deals with the development and use of DISTRO-II, a ma or modifiation of a previous BESRL model, the General Matrix Manipulator. DISTRO-II, whih simulates additional fators affeting personnel distribution, an assist management by more aurately prediting the effets of seleted poliy-related nondeployability fators. JHLANER Tehnial Diretor

8 DISTRO II DISTRIBUTION ROTATION MODEL BRIEF Requirement; To develop a model enabling Army personnel management to assess quantitatively the impat of personnel poliies on deployment and readiness of personnel. Researh Produt. The DISTRO-II model, essentially a modifiation of the General Matrix Manipulator for use in distribution problems, was developed to alulate deployability harateristis of the Army through simulation of priority groups and distribution areas, and the differential deployability of personnel in three ategories: 1) NOS not deployable overseas available for ONUS assignment only; 2) NST not deployable to short tour available for ONUS and long tour assignments; and 3) IAV not restrited generally available for assignment The DISTRO-II model expands the apability of the General Matrix Manipulator in the following ways: 1. Broadens the model's onept of poin-y aused nondeployability 2. Relates the world-wide availability of per«onnel to tour area manning levels 3. Extends simulation to priority groups 4. Fills tour area requirements in advane t I/'

9 DISTRO ll-distribution ROTATION MODEL ONTENTS BAKGROUND 1 The Deployability Problem 1 REQUIREMENT FOR DISTRO II MODEL 2 DISTRO II MODEL 3 TEHNIAL SUPPLEMENT 9 PRINIPLES OF THE GENERAL MATRIX MANIPULATOR 11 INTERFAE OF GMM WITH DISTRO-II 11 Step 1 Input DISTRO II Data 16 Step 2. alulate Personnel Assets 21 Step 3. Summarize Authorizations 22 Step 4. alulate Personnel Availability Groups 22 Step 5 Distribute Asseu and Available Personnel to Priority Groups 22 Step 6. onvert Priority Group Manning Levels to Tour Area Data 28 Step 7 Store DISTRO II Data 33 Step 8 Output DISTRO II Summarized Data 33 APPENDIX 39 DISTRIBUTION 53 DD FORM 1473 (Doument ontrol Data (R&D) 55 Page TABLES Table 1 Input format for DISTRO-II 19 t' Table 2 Variables input by user to DISTRO II )jt I 20

10 FIGURES Page Figure 1 Types of matries 4 2 GMMDISTRO II deployabihty ategories 4 3 GMMDISTRO II extension into priority groups 6 4 Most models fill tour areas like this 7 5 DISTRO II fills tour areas like this 7 6. Types of nodes in the GMM Types of node flows in the GMM 13 8 General matrix manipulator (GMM) logi 14 9 Interfae of GMMDISTRO II logi Entry INPUTD inputs DISTRO-II data Data setup for DISTRO-II Types of personnel within matrix Entry ASSET sums assets Entry AUTHOR sums authorizations Entry AVAIL summarizes available personnel Entry DISTR ontrols personnel distribution Subroutine ALLOT distributes personnel Entry GMMANLEV determines tour requirements Entry DATASTOR stores DISTRO-II data Entry OUTPUTD outputs DISTR02 summary 35

11 DISTRO ll-distribution ROTATION MODEL The Deployability Probler < BAKGROUND Projeting the Army's distribution apabilities its ability to provide qualified men lor speifi jobs at an appropriate time and loation is a diffiult task for management For many years. Army management has estimated its distribution apabilitie. primarily as a funtion of twu variables 1) the number of job vaanies and 2) the number f men qualified to fill these vaanies At best, estimates based on these data are rude a oximations of reality Many other fators restrit the deployment of qualified men to th» p& liular jobs where they are needed As long as the Army losely resembles a steady state system, with losses equaling replaements and authorizations remaining relatively onstant, disount fators an approximate the effets of inhibiting or nondeployabihty fators and the estimates are within aeptable limits However, when drasti dynami hanges suh as the buildup and subsequent phase-down of fores in Vietnam begin to our in the military setting, the rippling effets of poliy aused nondeployabihty skyroket It beomes very diffiult to estimate aurately the impat of rapid turnover and frequent poliy hange on personnel deployment and readiness Estimation methods whih take into aount more of the distribution restraints are needed Personnel deployment and readiness in the Army, as in other personnel systems, are a funtion of two major fators 1) requirements relating system onstraints and poliy onsiderations, and 2) personnel resoures refleting harateristis of individuals in the system. These requirements and resoures of'en appear well-balaned, whereas in reality several variables suh as rotation poliies and tour area needs may interat to redue drastially the pool of deployable personnel. The problem for management is basially one of mathing available resoures with system requirements while allowing all r «levant onstraints to interat realistially. Any hange in the nature of military onflit may ause differential losses due to deaths, injuries, prisoners of war, or retirements In turn,both position vaanies and the number of men qualified to fill the vaanies are affeted Rapid turnover aused by limiting ombat servie to a few months or by requiring only two years' involuntary military servie arouses major ost onsiderations relating to training and travel between assignments Management's values may also affet the nondeployabihty of personnel Management sets priorities inluding whih assignments should be stabilised for system effiieny, whih vaanies should be filled first, or whih poliies are the most expedient After suh deisions have been made, harateristis of the individuals in the system may still alter the piture onsiderably It may be too ostly to assign men with short expeted termination of servie (ETS) dates to new positions Available personnel may not possess appropriate skills and therefore have tobe letramed. In order to ontinue high morale among troops, individual rights must also be guarded, and onsistent preditable poliies arried out with respet to limiting the number of ombat area assignments, minimizing time in a ombat zone, or allowing aompanied tours between ombat assignments All these fators interat ausing tension between requuemenls and resoures.

12 Although rompuler mod«'ls pri>viuusly developed by the Behavior and Systems Researh Laboratory (BESRL) are not speifially designed to handle nondeployabihty, they use several tehniques whih partially reflet the deployability apabilities of the Army One of the earliest tehniques used in th«mass flow models disounted personnel inventories by a fator representing transients, patients, prisoners, and students (IPS) Later models approah the problem moie diretly by representing speifi types of nondeployabihty For example, some models maintain separate nodes for stabilized assignments, ombat tour returnees, and other speial ategories of personnel Simulation models may also represent delays between system entry and assignment as well as monitor estimated termination of servie Systems analysts emphasize areful design of assignment rules to faithfully inorporate system onstraints All these methods are helpful, but management needs a more omprehensive evaluation of the nondeployabihty problem It needs to meaningfully relate management poliies, system variables, and individual harateristis so that more nondeployabihty fators an be aounted for REQUIREMENT FOR DISTRO II MODEL The Program to Improve Management of the Army Resoures (PRIMARI Projet 5-1. Developing Tehniques for Assessing the Impat of Personnel Poliies on Deployability, expressed management's onern about non deployability An expliit goal of this projet was the development of proedures for quantitatively assessing the impat of personnel poliies on deployment and readiness A speial team of the SIMPO I Steering ommittee was assigned to monitor an adaptation of the General Matrix Manipulator (GMM) to over poliy-aused nondeployabihty ' The GMM was seleted beause it is a matrix-based omputer model whih traks both time in the system and time in the tour for all personnel Also, reassignment flow rules an easily be input and hanged by the user at the time of exeution The two time dimensions, as well as the rotation rules, are important fators affeting the deployability of personnel Staff Offiers of the apabilities and Analysis Division. Diretorate of Prourement and Distribution. Offie of the Deputy hief of Staff for Personnel (DSPER AD), agreed to work with BESRL in developing a problem in whih interations between rotation and assignment proesses and other nondeployabihty fators ould be examined BESRL s initial response to the PRIMÄR requirement was the SIMPO I Distribution-Rotation Model (DISTRO I), an adaptation of the GMM to the deployability problem. 2 Within the original GMM framework. DISTRO I subroutines distribute personnel in the GMM tour areas to ommand elements At the end of the simulation, all nontransients within a given lour area, following disounting for TPS, are distributed to ommand elements proportionately to ommand element authorizations Although DISTRO I represents nondeployabihty more aurately than earlier models, it has some limitations The system's distribution apabilities are not diretly related to tour area manning levels during the simulation The total number of available personnel serves only as a eiling in determining tour area manmn levels DISTRO I does not onsider the wond wide availability of personnel, it distributes eah tour area separately without relating it to the total system Separate distribution of eah tour area to ommand elements is, at best, a fragmented approah ' Witt. Joanne M. and Adele P Narva. SIMPO I General Matrix Manipulator (GMM). Tehnial Researh Rtport 1165 Behavior and Systems Researh Laboratory, Arlington. VA. Jan'i.?rv Witt. Joanne M. SIMPO I Di-TRQ Distribufio, Rotütion Model. Researh Memorandum 714. Behavio and Systems Researh LatWütory, Vlmgton. VA. September 1971 o.

13 DISTRO II MODEL To more aurately model deployabihty harateristis of the Army, BESRL sientists with the guidane of DSPER personnel have developed the DISTRO II Model, a set of subroutine options whih signifiantly modify the GMM. DISTRO II routines alulate deployabihty indies whih interat with the assignment-rotation proess throughout the GMM DISTRO M routines have four major effets on the GMM simulation 1 They broaden the model's onept of poliy-aused nondeployabihty 2 They relate the worldwide availability of personnel to tour area manning levels 3 They extend the GMM simulation to priority groups 4 They fill tour ar 3 requirements in advane DISTRO II broadens the definition of deployabihty in that it aounts for more nondeployabihty fators than does either the GMM alone or DISTRO I (Figure 1) Personnel matries in the GMM alone represent all personnel as assets. DISTRO I separates nndeployable transients from deployable assets Based on time in tour, DISTRO II matries represent three kinds of personnel transients, assets, and reassignables, eah with its own deployabihty harateristis The transients in Figure 1 are those who are reassigned to the area but who are not yet onsidered assets Assets are those who art? atually in an area anu who annot be reassigned, whereas the reassignables are available for reassignment DISTRO II employs the onept of differential deployabihty Differential deployabihty means that a person is deployable with respet to ü speifi area, eg., an individual may be deployable to long tour and not to short tour, or to ontinental United States (ONUS) and not to long tour In other words, deployabihty has meaning only in referene to speifi area needs In DISTRO II, differential deployabihty is represented by three ategories of personnel availability: 1 NOS not deployable overseas available for ONUS assignments only 2 NST not deployable to short tour available for ONUS and long tour assignments 3. IAV not restrited generally available for assignment These availability ategories reflet restritions plaed on individuals available for reassignment As eah man ompletes a tour or otherwise beomes available for reassignment, he automatially inreases the tally in one of the three availability indies (Figure 2). For example, a few men are stationed in ONUS for reurring tours beause of speial personal irumstanes or beause of their unique skills As eah of these men ompletes a ONUS tour, he adds one to the NOS ounter. Most short tour returnees are deployable only to aompanied ONUS and long tour; these men are ounted by the NST variable. Others, ompleting two years in ONUS or the long tour and being eligible for any assignment, are added to the IAV ategory. The sum of NOS, NST, and IAV equals all personnel available for reassignment at the end of a time period. The three ategories enompass most of the restritions on deployment.

14 GMM ALONE GMM-DISTRO II TIME IN SYSTEM 1 TIME IN SYSTEM o 3 O K Z Ui Z ASSETS o t- z UI TRANSIENTS ASSETS h- REASSI6NABLES Figure 1, Types of matries TIME IN SYSTEM D 3 O H Z UJ 1 TIME IN SY STEM NOT OVERSEAS REASSIGNABLES NOT SHORT TOUR GENERAL 1 Figure 2. GMM DISTRO II deployability ategories - 4 -

15 In the GMM, input to a given tour area is a diret funtion of that tour area s authorization without onsideration of the availability of personnel assets By applying the DISTRO I options, tour area manning levels are made to reflet the total number available for reassignment These manning levels, however, ignore differential deployabiiity by assuming that deployable personnel assets an be assigned to any area Employing DISTRO II options, manning levels in the GMM reflet authorizations, assets, and the number of men available for reassignment to speifi areas To aomplish the best distribution of available personnel to tour areas. DISTRO II first distributes the most restrited group of personnel, NOS, to ONUS Following the NOS assignments, NST personnel are distributed to ONUS and long tour Then personnel in the least restrited group, IAV, are distributed among ONUS, long tour, and short tour This approah prevents f illmg one tour area to its fully authorized strength only to find that the last tour to be filled has a ritial shortage. Sine the model initially onsiders speifi availability of personnel in determining manning levels, all tour areas are filled equitably Present Army management proedures for utilizing available assets are based on five distribution groups, alled priority groups, whih are filled relative to their importane to mamtaimngamaximallyeffiient military system The GMM works to fill tour area authorizations wi.hout onsidering priority groups Employing the DISTRO I routines, tour area personnel are distributed at the end of the simulation into the five priority groups The priority groups, however, are not onsidered during the simulation A new approah was needed to distribute personnel to eah of these priority groups aross the entire system not separately within eah tour area Instead of determining how many are needed in the short tour and then saying that N of these will be in priority group 1 as m DISTRO I, management would like to determine how many are available for priority group 1 and then send N of these to short tour. The DISTRO II proedure plaes major emphasis on filling priority groups aross tour areas rather than filling speifi tour areas. The ommon element linking the GMM simulation matries to the expanded simulation of priority groups in DISTRO II is the tour area DISTRO II totals assets within individual tour areas and distributes these assets into five priority groups whih are not represented diretly by the GMM matries (Figure 3) DISTRO II subroutines determine how many of the total tour area assets are serving in eah priority group within the tour area and how many of the available assets will be assigned to eah priority group aross all tour areas onsidering their urrent assets and authorizations Tour area manning levels are then realulated by apportioning new personnel m eah priority group to the individual tour areas These tour area manning levels are used as a driving fore in the GMM to reassign personnel Thus, the distribution proess beomes an integral part of the rotation-assignment proess within GMM-DISTRO II Figure 4 shows how most models, inluding the GMM, fill requirements The top priority areas get 100% or maximum fill, the middle priority groups get less, and the low priority areas get what is left over Often this method results in a ritial shortage in one or more areas. It is neessary to maintain a minimum number of men in low priority areas in order that high priority areas an funtion adequately. Employing the DISTRO II options of determining manning levels based on the total world-wide availability of personnel, all areas an be filled adequately in relation to the total system The emphasis shifts from tour areas to priority groups, and Figure 5 shows the results The priority groups are filled in relation to their importane as speified by management. DISTRO II also direts the GMM to fill requirements one month in advane to allow suffiient travel time for men to reah their destinations. Manning levels for month two are atually filled in month one The model determines manning levels for month two based on the number of projeted assets for month two and the number available for reassignment in month 1. DISTRO II, employing a ombination of assignments in advane, differential deployabiiity, and simulation of

16 6MM TOUR AREAS 6MM-DISTR0 II PRIORITY GROUPS 6MM TOUR AREAS Figure 3. GMM-DISTRO II extension into priority groups

17 m to eg r- -wr lit I»I I f (A Z O o ^ to (Si * ' >' i" i! OQ 23 OO _!»- «# /M TT -H-H /\ 3 Hr* (,E7 4* fetri ; *i. Min.» i Uli"» -.'.' ( m&m n Ul < o Ü < o o O X 0) -K «3 w JZ ** 0) ^!2 (0 e «3 o_ 3 o o X p (n H»^ 8 5 0) i o SSI fevymmi ''ärm»i>.^i».'iia'..v t /SlMi :T 1 T L! BEI 1 Tlld in30w3d 01 TO (I) 3 O in = o o I- n Q in I - 7 -

18 priority qroups, oupled with reassi()nim,'nt poliies as represented in the GMM inludim) separate representation of different types of deployable personnel- presents a more realisti piture of poliy aused deployability for manaqement than did pievius models. DISTRO II an I«? useful to management in problems suh as the followim): 1. How an management provide more experiened personnel in the short tour? If extended tours are allowed in return for various alternative rotat on pol.ies, how many more experiened personnel will be available? At what point in time would the alternative poliies beome infeasible? If manatjement promotes men faster, at what point does the system have a shortage of men with the minimum qualifiations for promotion? 2. What are the best poliy alternatives whih will limit the number of ombat tours for areer personnel? 3. What are the effets of allowimi more time between assiqnrmnts? These and many other questions an be studied with the use of GMM DISTRO II so that rnanaqement an better predit the effets of nondeployability fators.

19 DISTRO II DISTRIBUTION ROTATION MODEL TEHNIAL SUPPLEMENT: PROGRAM DOUMENTATION

20 PRINIPLES OF THE GENERAL MATRIX MANIPULATOR The GMM is a matrix based mass flow model whih simulates rotation poliies, skill aquisition, and assignmentpoliiesof Army management Groups of personnel are represented by numbers arranged in two dimensional matries suh as those in Figure 6 The dimensions may be defined by management as time in servie and time in tour or time m servie and time in grade These fators are important for determining the deployability of personnel Eah mat; ix representsagroupof similar individuals, suh as 11B1 in the short tour or 11B? intheloigtour Movements between these matries are determined by management poliit-s These movements represent poliies of two types rotation and promotion (Figure 7) Type 1 flow, simulating promotions, is aross skill levels a hoi i/ontal move in whih a person s nme in tour remains the same while his time in grade moves!o the initial position of the next grade Type 2 simulates reassignments Movement between ommands is from one time in grade and time in tour to the same time m grade and the first time position in another tour Reassignments an our at the end of a tour of duty or prior to the end of a tour of duty Employing these matries and flows, the GMM logi is relatively simple (Fu^ 're 8) It updates the matries, alulates assets, determines requirements, uses flow rules to fill these requirements, and then repeats the yle for eah subsequent time period INTERFAE OF GMM WITH DISTRO II DISTRO II options are integrated with the GMM simulation proess Under these options, DISTRO II subroutines gain ontrol to shift the emphasis of the simulation from tour areas to priority groups spanning the tour areas To aomplish this, the DISTRO II subroutines onvert the GMM tour area data to priority group data The priority group data are manipulated so as to represent distribution proesses in the system They are then reonverted to tour area data As a onsequene, the new tour area data reflet priority group hierarhies and yet are still in a form ompatible with the GMM The GMM then regains ontrol and ontinues to operate on the tour area data as if it had not been altered By swithing bak and forth between GMM and DISTRO II routines, the simulation of matries representing tour areas and personnel ategories is expanded to over a new dimension of priority groups This expansion ours with minimal inrease in online storage and data preparation If these priority groups were simulated diretly, the number of matries in the GMM would have to be multiplied by fiv» Dne matrix for eah priority group would be needed to replae eah personnel matrix For example, instead of one matrix to represent personnel on their first short tour, the GMM alone would require five matries, one for personnel on their first short tour in eah priority group Besides making the data base massive, :low rules among the matries would beome muh more ompliated and repetitive DISTRO II an be expressed as a series of events interwoven with the GMM Figure 9 illustrates [his integration of GMM and DISTRO II logial events DISTRO II events are summanh d in the following eight steps 1 Input DISTRO II data 2 alulate personnel assets 3 Summarize authorizations by distribution or tour areas 11 Preeding page blank

21 D UJ 0- > lo UJ-J -> l-ui J MDOi Nl 3Wli Q- 4 UJ Ü 2 >? (E (9 UJ v «n x: 2 ^ UJ ^m * -»- o 1 ^ H» > h dooi Nl 3WI1 I isi - 12

22 Z QNVWWOO I QNVWWOO T

23 START) INPUT PARAMETERS INPUT SYSTEM DATA START TIME PERIOD 0 ALULATE NODE ASSETS ALULATE NODE REQUIREMENTS! FILL NODE REQUIREMENTS TIME PERIOD» TIME PERIOD + I MAKE FINAL TRANSFERS UPDATE SYSTEM INPUT ADDITIONAL PERSONNEL MAKE INITIAL TRANSFERS STOP Fi( ure 8. General mdtnx mampuldtur (GMM) loqi.

24 u 'Öl o O E t- ( ) Q i O» u 15

25 4 alulate personnel availability groups b Distribute assets and available personnel to priority groups 6 onvert priority group manning levels to tour area data 7 Store DISTRO II data 8 Output DISTRO II summan/ed data Step 1 is exeuted one at the beginning of the simulation Steps 2 through 7 are exeuted in sequene one during eah time period of the simulation At the end of the simulation, Step 8 is exeuted These steps, monitored by the GMM mam program, MAINGMM, are exeuted by eight separate entry points in the mam DISTRO II subroutine, DISTR02, and by an auxiliary subroutine, ALLOT (FORTRAN Listings of DISTR02 and ALLOT are in the Appendix) The following setions desribe these eight steps in detail Inluded are flow harts and desriptions of the omputer routines to aomplish eah step and the way in whih the DISTRO II routines interat with those in the GMM Along with other GMM parameters, the variable IDISTON is input at the beginning of eah GMM simulation If IDISTON is equal to or less than zero, the GMM proeeds without DISTRO II intervention If IDISTON is greater than zero, however, the following eight DISTRO II steps are exeuted as they our in the GMM simulation Step 1 Input DISTRO II data The GMM relinquishes ontrol to DISTR02 Entry INPUTD (Figure 10) whih inputs DISTRO II data The data define the relationship of matries representing tour areas within the GMM to priority groups (PG) within DISTRO II. Input data setup, format, and variable definitions are desribed in Figure 11, Table 1. and Table 2 Entry INPUTD reads four different types of data: major parameters, availability parameters, asset parameters, and authorization and manning level parameters Boundaries for the entire DISTRO II simulation are set by the major parameter ard The remaining input parameters operationally define these boundaries For example, the number of availability ategories, NAVAT, input in the major parameter ard, is defined in the availability parameter ards by the vanablespernos, PERNST,and PERIAV These variables are perentages of personnel in eah matrix who are available for the three different types of assignment NOS, NST, and IAV Asset parameters identify GMM matries by distribution areas for use in alulating assets Authorizations are input for priority groups within eah distribution area For example, if there are three distribution areas (DAI, DA2, and DA3) representing short tour, long tour, and ONUS respetively, and if there are five priority groups, then there will be 15 authorizations one for sh}rt tour priority group 1, one for short tour priority group 2, et Authorization and manning level parameters also inlude fill rates, whih indiate the relative importane of filling eah of the five priority groups aross all distribution areas. In order to represent only one priority group within a given distribution or tour area, the user would input zero authorizations for the other four priority groups within the area. The remaining seven steps are totally dependent upon the setup of DISTRO II parameters These parameters must be aurately estimated in order to meaningfully represent distribution apabilities of the system Following the input of DISTRO II data, GMM regains ontrol to input the remaining GMM system data and to begin the simulation. After updating the system and making initial personnel transfers, the GMM teams up with 0ISTR02 Entry ASSET to take an inventory of personnel assets

26 f ENTRY 'N V^ INPUTD J,'^1NPLT & PRINT j MAJOR PARAMETER ARD Ü INPUT & PRINT AVAILABILITY PARAMETERS \L INPUT & PRINT ASSET PARAMETERS.N INPUT & PRINT AUTHORIZATION & MANNING LEVEL PARAMETERS E (RETURN) Figure 10. Entry INPUTD inputs DISTRO 1! data - 17

27 AUTHORIZATION AND MANNING LEVEL PARAMETERS BSBSflMffiL 71 GROUP IDENTITY DISTRIBUTION GROUP IDENTITY fj INPUT AVAIIABILTTY 71 ASSET PARAMETERS AVAILABILITY PERENTAGES AVAILABILITY PARAMETERS MAJOR PARAMETERS Figure 11. Data setup for DISTROII

28 Table 1 INPUT FORMAT FOR DISTRO II ard Desription No. of ard* Variable Name olumns Format Major Parameter ard NODISGR NPERAT NOPRGRP no NMATRIX no NAVAT no Availability Pa i- meter ards (NMATRIX* NAVAT+8- NAVATI/8 PERNOS(I) PERNST(I) 1-10 F F103 PERIAV(I) (1=1,NMATRIX) F10.3 IOS/8 losdgr(l) (I I.IOS) Asset Parameter ards NMATRIX 8 MATWDGR(I) (1= 1,NMATRIX) NMATRIX/8 MATWPER(I) (I i 1,NMATRIX) 1 10 no Authorization and Manning Level Parameter ards (NOPRGRP* NODISGR) 8 (NODISGR* NPERAT* NOPRGRP)/8 LAST/8 R(l) 1-10 F103 (1^ 1,N0PRGRP*N0DISGR) AUTH(I) 1-10 no (I 1.NODISGR*NPERAT*NOPRGRP) NEWAUTH(I) 1-10 HO (I FIRST.LAST) - 19

29 Table 2 VARIABLES INPUT BY USER TO DISTRO II Variable Definition Maior Parameters NODISGR NPERAT NOPRGRP NMATRIX NAVAT Number of distribution groups or tour areas Number of personnel ategories within eah tour area Number of priority groups withn eah tour area Number of matries with the GMM Number of availability ategories Availability Parameters PERNOS(I) PERNST(I) PERIAV(I) losdgr(l) Perent of personnel ompleting matrix I who are eligible for ONUS only (not overseas) (Distribution Group 3) Perent of personnel ompleting matrix I who are eligible for ONUS and LT (not short tour). (Distribution Groups 2-3) Perent of personnel ompleting matrix I who are eligible for ONUS, LT, and ST (Distribution Groups 1, 2 and 3) Distribution group or tour area for Ith input ategory Asset Parameters MATWOGR(I) Distribution group of tour area for Ith matrix MATWPER(I) Personnel ategory for Ith matrix Authorization and Manning Levoi Parameters R(l) Fill rate for Ith distribution priority group AUTH(I) Authorization for Ith distribution priority group NEWAUTH(I) Vetor ontrol for inputting new authorizations for month I If in month I the NEWAUTH(I) equals 1, new authorizations are read in If the NEWAUTH(I) equals zero, the authorizations previously read in are used.?0 -

30 Step 2. alulate personnel assets. At any point in time, eah matrix an be represented as a ombination of transients, urrent assets, and reassignables (Figure 12). TIME IN THE SYSTEM 12 MAXLEN 1 1 M E 1 r > i T 0 U l R N L- L L* 1 TRANSIENTS (TR) ^OMPLETIONS REASSIGNABLES (RE) "I H "l I I I I i _Assets for following TPi Assets for urrent TP 0 F igute 12. Types of peisunnel willun matux GMM subroutine SUMMARY alulates matrix assets for the period (TP 1 ) following the one urrently being simulated (TPQ). Eah matrix IS summed from row 1 in the tour through row L-1 in the tour and olumns 1 through MAXLEN in the system. Matrix assets (AS) for TP 1 are equal to the sum of the transients and assets for TP 0, exluding those in their final month in the tour (ompletions) AS T p - TRANSIENTS-rp Tf 1 TP 0 ASSETS TPr OMPLETIONS TPr This projetion of assets for the following TP 1 assumes that transients in the urrent TPQ will beome assets and that personnel in their last month in the tour (ompletions) will beome eligible for reassignment in the following TP 1. After assets have been determined for eah matrix. Entry ASSET (Figure 13) totals the assets for all matries within eah of the three distribution areas (DAI. DA2, DA3) AS DAI Assets for DA1. AS DA2 Assets for DA2. AS Assets for DAS DA3 The objetive of the remaining steps is to input men into the proper DA and PG one TP in advane After updating at the end of the urrent TPQ, the urret number of men should be serving in the areas where they are needed -?1

31 Step 3 Summame authomations The Ith element of the NEWAUTH vetor is heked during eah time period If the element equals zero, the last authorizations input to the model are employed If the element is positive, DISTR02 reads new authorizations for eah PG and DA Entry AUTHOR (Figure 14) then sums these authorizations for eah DA AUTH Authorization for Ith PG within Jth DA DAJ,PGI The GMM regains ontrol to input new personnel into the system after whih ontrol returns to DISTR02 for alulation of personnel available for reassignment. Step 4 alulate personnel availability groups. In eah matrix during the TPQ- all personnel available for reassignment an be broken into three ategories: (1) NOS, (2) NST, and (3) IAV Entry AVAIL (Figure 15) multiplies the number o* reassignable personnel (RE) from eah of the matries in the GMM by its availability parameters (PERNOS, PERNST, and PERIAV) and determines how many people are ategorized as NOS, NST. and IAV To these totals it adds new system input to the appropriate eligibility ategory. NOS E (RE, * PERNOS,) + INPUT N0S NST J (RE, * PERNST,) + INPUT NST IAV S (RE, * PERIAV,) + INPUT AV Steps 1 through 4 are not an integral part of the distribution proess. They onvert GMM data into a form ompatible with DISTR02 and then alulate summary data. This data preparation is neessary prior to the atual distribution proess, whih is aomplished by Step 5. Step 5 Distribute assets and available personnel to priority groups. The distribution proess onsists of a series of separate distributions of different groups of personnel. The entire proess is direted by Entry DISTR (Figure 16), the brain of the DISTRO II model. Aording to the instrutions given by Entry DISTR, Subroutine ALLOT aomplishes the atual distribution of speifi groups of personnel. Following eah separate distribution by ALLOT, ontrol returns to Entry DISTR for additional instrutions. Initially, DISTR direts ALLOT to distribute urrent assets in eah DA to PG's within that DA. These assets will have an influene on the number of additional personnel assigned to eah 0A- PG. After the urrent assets have been distributed within their own DA's, DISTR direts A L LOT to distribute all personnel available for reassignment to PG's aross appropriate DA's onsidering urrent assets in the DA-PG, authorizations for the DA-PG, and thn number of personnel available for that DA-PG. First, the most restrited personnel, NOS, are distributed to DA3 Then NST personnel are distributed among PG's within DA2 and DA3 Finally, the most unrestrited personnel, IAV, are distributed to all PG's within all DA's. This sequene of distributions is desribed below in detail, inluding the atual steps whih Subroutine ALLOT takes to aomplish eah of the distributions. -?2

32 KNTRY "N ASSKT J SET DGRAS8ET VP:TOR AND PRASSET Mi\TRIX = 0 J - MATWDGR(l) DGRASSET(J) = DGRASSET(J) + ATUAL(I) k SUM ASSETS BY DISTRIBUTION GROUPS, OLLAPSING AROSS PERSONNEL ATEGORIES ASSETS(I) ATUAL(I) I + 1 RETURN Figure 13. Entry ASSET sums assets 25

33 GENTRY 'S AUTHOR^ SET PARAMETERS FOR READING AUTHORIZATIONS SUBROUTINE PREPARE INPUTS AND TOTAI5 AUTHORIZATIONS ALL -A SUBROUTINE PREPARE ( RETURN J 14. Entry AUTHOR sums authorizations H -

34 To transfer ontrol to Subroutine ALLOT, Entry DISTR uses the following all statement all ALLOT(NEWIP, IDIS, NOPRGRP, NODIS, IOVER) The values of the input elements NEWIP, IDIS, NOPRGRP, and NODIS determine the following onditions for the distribution proess, alulated in DISTR, NEWIP amount of new input available for distribution IDIS first DA to whih NEWIP will be distributed NOPRGRP - number of PG's within eah DA NODIS - number of DAs aross whih NEWIP will be distributed The element, IOVER, is alulated in ALLOT and is returned to DISTR as the number of available personnel over and above PG authorizations This overage will be distributed to DA3 at the end of the distribution proess Initially DISTR direts ALLOT to distribute urrent assets within DAs to PG s all ALLOT(AS L:)A.], NOPRGRP, 1,0) Subroutine ALLOT gams ontrol and proeeds with the following steps to aomplish the distribution (Figure 17). a. It obtains asset totals for eah PG aross all DA's. IDIS+NODIS-1 SUMASS(I) E (AS DAj> p GJ ) pidis If ALLOT is distributing urrent assets, the assets (AS ) DAj.PGi are equal to zero prior to the distribution Therefore, SUMASS(I) equals zero b. It obtains authorization totals for eah priority qroup aross all DA's onerned. NODIS+IDIS-1 SUMAUTH(I) - E (AUTH DAj( p Gl ) jhdis. It obtains priority group weighted shortfalls aross all DA's. PRNEED(I) - (SUMAUTH(I) - SUMASS(I)) * R(l) Also, it obtains the total weighted PG shortfall NOPRGRP SUMNEED Z {PRNEED(I)) 1 =

35 Figure 15. Entry AVAIL summarizes available personnel 26

36 mir MMn (IF ün TO nr DISTKIHITI» 1 ALL si likoutink ' AI.IOT TO D1S1RI-I. Mm l)vkra(.es 101 i LAST nmthiuj I URmil' NJ.Uir " NST I - I t I ~NtM\ F'~ HIMlRd) J m 1 NI.W 11' M-:w 11' + NOS i AU. SUBHOUTim: I AI.IXJT Tu D1STK1-I BITE NKWll' FOR I I th UISTRIBITIONI RÜli' I AM, SUBROUTINE I I ALUIT TO DISTRI-I I BITK NEW1P TO I 1 th DISTRIBUTION! J ROUP Figure 16. Entry DISTR ontrols personnel distribution 27 -

37 d Using the information obtained in steps a through, it determines h )w many personnel should be input to PGs within DAs e NEWIN -- (NEWIP SUMNEED) * PRNEED(I) ' (AUTH 'SUMAUTH(I)) OAi.PGi DAj.PGi It adds the new input to the respetive DA P assets ^DAj.PGi ASQAJ.PGI + NEWIN DAj.PG Authorizations for DA PG serve as eilings for DA P exept when urrent assets are being distributed within their own distribution areas When distributing NOS, IAV, and NST personnel, DA2 and DA1 should be held to their authorization levels All exess personnel (those over the authorizations) should be aumulated in the variable lover f After urrent assets have been distributed within their DAs, DISTR dire-rt- AJ OT to distribute NOS men to DA3 all ALLOT (NOS, 3, 5, 1,0) Employing these parameters. ALLOT repeats steps a through e g ALLOT is d reeled to distribute NST men to DA2 and DA3 all ALLOT (NST. 2, 5, 2. 0) (Repeat steps a through e) If IOVER is greater than zero, I0VER is distributed to DA3 h IAV personnel are distributed to DAI. DA2. and DA3 all ALLOT (IAV, 1, 5, 3. 0) (Repeat steps a through e) i If there are men above the authorization levels in DA2 and DAI, (IOVER), distribute IOVER to DA3 all ALLOT (IOVER, 3,5, 1,0) After all personnel have been distributed, the matrix AS represents projeted DAj. PGi manning levels for the Ith PG within the Jth DA for the following time period (TP,). Step 6 onvert priority group manning levels to tour area data Entry GMMANLEV (Figure 18) totals PG manning levels within eah DA NOPRGRP MANLEV DAJ v AS DAJ p Gi II - :> -

38 f STAKT J JSIAKI IPIS *N()IK(;RI'. uiis mis * NOUIS - 1 I. ISTAHI - 1 PRW:H)(K) 0 nor IAI'TH i SIMMHI 0 TWASS \n.am K 0 K. K I o SlWffH) PVNFKIKK) IDIS - 1 sinann, SlIMAIR, S1IMASS, AND -HtKtl) VT' 4IS J - J + I itlmalth(k)-suhaitrh(k) I'RALTIKr.J) il-malr(k) SUMAUX(K) PHAirrHd, i)«i((k) SIWSS'fK) mi SIINASS(K) 4 PRASSKT(K, I) TrrASS(J) TOTASS(J) fprasskt(k,j) PIIIED(K) (SUHAITH{K) - SUIASS(K)) * MQ AIfl'l/TF ASSIT ANII AinilKI/AI ll/. TOTAI-S Fl* Kl'i PKIOKITY UKOIf AflKOSS DlSTKimTION HOtlS KTF.KHI» mourn KOUP SH(*IA(:KS I - I «I ITOT ITOT T(rrASS( n 1AUTH lalth + DKAUTIUJ) X NEWIP RATIOX/SIIMftKI) MKW 0.0 J IDIS I BT( IN DISTRIBI'TION OF «TWIP AHüN<. K PRIOKITV IKOI'PS. TI»;N DISTKI- BITT KAtM PRIORITY KOI'P TO DISTR i.kps Figure 17. Subroutine ALLOT distributes personnel r«.

39 I 9 I.. I 1 IMAKI.IN 0 I, I «HKt.RI'- mmtf + i i'k - imvnfftj) WU[N(I )"KATIO- v twhkhw PH/SA IX.KAI IX.RAUTIK.I) D.RAS MWII' 'LIMITED «o sit AND»Wlf -fs ^ nrrr. laiti I i m N»:UIN(I.) - (IK/IM.KA! ) 1M.RAS *- xrawin-nt^inii.j HH.IN KODSDIW KKHOK am t NMHV(t) - i, RASsrT(r,.n StUINd.) ilk.f NVW " MANN1NI. mill A( AIS'ST i AunotiurioH N>.W-NVWIN(I.) HAMIiVd) ikainiidi, o N1UINII 1 MANU.V<I.). IHASSKKK, I) TRASSIKK. V ^ W WIN(1.> 0 (ianiivd.) r(assit(k, I) & Figure 17 ontinued 50 -

40 1 X X NJUINd ) SIW» HI MTW.KIj -((HWinHK,.!) l, IIASSIT(l(..l)*R(Ki iiinmm*(i) SIIMAUT'UK). inarnuk, i) Mill X/ SI'MWKI) PRASSKKX, I) MANIiVd.) I IS1ART 1 J IDIS - 1 DF.TKRMINt NimKR OK EXWSS IfKSllNNH. J - J+1 K I K K + I I - I + 1 N>.U»,UIN(I) m:wn' NKM-NHJ Figure 17 ontinued 7 1

41 hrnma pittiin i n i KM 1'»n.'i'rKiMi im Figure 18. Entry GMMANLEV determines tour requirements - 3?

42 These DA manning levels will serve as tour area manning levels in the GMM From the DA manning levels, Entry GMMANLEV subtrats the atual number of urrent assets in the tour area to obtain the number of men who will be input to the tour areas by the GMM NEEDS DA j MANLIV )AJ ASSLTS naj The GMM then regains ontrol and fills these tour area needs aording to the priority of fill rules Step 7 Store DISTRO II data At the end of the time period. Entry DATASTOR (Figure 19) gams ontrol, alulates summary data, and stores the data on a disk pak for use at the end of the simulation Entry DATASTOR totals assets for the time period by summing rows 2 through L m the tour and olumns 1 tin ough MAXLEN in the system in eah matrix It totals transients in row 1 and olumns 1 through MAXLEN for eah matrix also These totals are then arranged within tour areas and stored on the disk Personnel distributed by DISTR are projeted assets and transients for PGs within DAs These projetions determine manning levels in GMMANLEV whih are m turn filled aording to the priority of fill rules in the GMM The data stored in DATASTOR summarize the atual number of assets and transients for eah matrix in the GMM after personnel have been moved These atual assets and transients are the data whih are of interest to management Step 8 Output DISTRO II summariied data Entry OUTPUTD (Figure 20) initially reads the data stored by DATASTOR It then transforms the matrix data into DA data and redistributes it to the PGs aross DAs Transients from the urrent time period TPQ being output are distributed in relation to assets for the following TP^ OUTPUTD prints the following information for eah time period whih has been simulated TIME BLOK N DISTRIBUTION GROUP J Priority Number Manning Perent Group Authorized Level Fill Shortfall Transients This output desribes the time period whih is bi og summarized, the distribution area or tour area, and the priority group within the tour area Eah priority group is further desribed by its authorization, atual manning level, perent of authorizations filled, its shortfall, and the number of transients who will beome assets in the next time period This output presents management with a detailed piture of where its personnel are loated The DISTRO II output loates men in priority groups within distribution or tour areas By looking also at the GMM output, management an see the speifi types of personnel who are serving in any one distribution or tour area. V -

43 .DATASTOR SET ITRAN AND IATIVE VETORS - 0 I - 0 i ITRAN(J) - ITRAN(J) + ITRANS I - I + I SET VARIABLES ITRANS - IATUAL4 IREAD 0 IUSE " 1 SUBROUTINE SUM- EY READS I th I RIX, TOTALS ETS, AMD SETS ' IATUAL«ASSETS(I)I M IREAD - 1 ALUUTE ASSETS FOR EAH MATRIX YESiU- K (1NTEG*2)-1 I. ITE UTIV VETOR OF Kth REORD ON DISK (LU 14) STORE DATA SUMMARY ON DISK IUSE - 3,_Jr,---- I SUBROUTINE I SUMMARY TOTALS TRANSIENTS AMD I ' SETS ITRANS I I ITRAM(I). S U - (MATWDGR(I)* NPERAT)-NRRAT + MATWPER(I) IATIVE(J)-IATUAL + lative(j) ALUUTE TRANSIENTS FOR EAH MATRIX K - K + I WRITE ITRAN, VETOR AS Kth REORD{ ON DISK LU U) I RETURN Ö Figure 19. Entry DATASTOR stores DISTRO-II data - M

44 / ENTRV N L OIITPITD J i 1 H1 NT / / OUTPUT / / Tim / I BEIN RF.DISTRIKI'TION OF DATA WITH IIISI«. K 0 I - FIRST - 1 START AVTAI ION OF UlSTRI r ihut FOR EAP.hK PERIOD (TP) SIh.LhTKD M - M + I ASSKTS(J) - ASSETS (.1) + lative(m) ITRAN(MI RI', I ASKTS(J) - 0 SET ITRAN AND INPl'T VETORS KQUAI, TO ZERO th fm K r VKTOR FROM (DISK (1,1121) iiative.vetor IPRIO 1PRIO + MATSUMdPRIO) KK KK * I PRASSETS(KK,,I) 0 ZERO ÜU1 VEIORS PRIOR TO PRINTIN AI-HIATIONS BEIN DISTRIBUTION OF ASSETS FOR TP I Ni:WlP-ASSETS(.n IDIS " J uns,i+nodis-l NODIS 1 Ö Figure 20. Entry OUTPUTD outputs DISTR02 summary 7.

45 ntisniiun TI 1 I ASSHS WITHIN I1ISTR RP,1 in mokitv uun BltniBUTI 1RANSIKN1S K(IR IP 1-1 AS A I-1 NI ION Of IP I ASSHS tun ««IF in I ' pniokir. GMUN ' I I uns mis + NODIS 1 IK - II1IS - 1 ^ZH-- KK - KK + 1 l'ka(:tlw,(kk,.n I I'RASShKKK, I) SIOKK ASSK1S nistrlbited TO I'RIÜKITY ROUPS WITHIN UISTR. t;rp. jr KK KK + 1 IAXNOPRRP* IK)-NOIKRlfKK MATSim(IAT)" MATSKMdAT) fnp:win( K^T) ** INP - 1NP + 1 NEW1P»INPUT(INP) NODIS IJ Figure 20 ontinued - 3e

46 ^ mis TIITAI. IiISTKImto ASSK'. K)R III I IK IK + 1 I notlihioitri 'MIN NdlULKl"* NOD I Si.R + I i IMlhX MIN - I fat IK / XPRATII - MAITIUKK, IK) IHRA-VRAIKKK, IK' PERf ll.l - PRAI KKK, IK) /XlKAirril ISilORT-XPRAHII- flfw I'RAIl l(kk, IK) MATSIW(MIN) - NATItWffllN) IMUTNUE, IK) MIN MIN + I E±-'-- MATSIIM(MIN') KATSUM(MIN) + ITUnKKK, IK) MIN - HIN 4 I MATSUM(MIN) MATSUM{M1N) + MATSIIM(IAT) ROI I' DATA FOR DISTRIBUTION ROIIH J IN TPI ITOM IIISTR()2 TOTALS IN MATS I M VAKIAhlJ PRINT TOTALS (F* PRlüRITY Figure 20 ontinued /GROUP KK S

47 0 l'kai I (UK, IK) STORK DlSTRIhl'TK.I) ASSI IS Kl«I PI IN PRATI VARIAM* I'KA'I IVf:(KK, IK) + MA'ISIMI II Al 1 y is N(l I'K-M/MSI M(MlN-?> AS-MAISIM(MIN-l) IT-MHI.I^AS/I'U si3(ti K! + 1 ZEE^-- 'RI-AD K" ' VFTOR KKUM DISK fill?1 1 - MIAN VI i Hill RIAIl l" 1 ' VI rnir KRilM [DISK I M> 14) INIW VI loh "~T MAS ^, ''WS I Tiw\ l'l RIOO KEN din in. KKAI) IKANSIINIS AND AVAIUUIU KOK XT I I SHORT - NATSlIM(MIN-2) ^~ MAISlMMIN-l) Figure 20 ontinued 3 P -

48 APPENDIX FORTRAN LISTINGS OF DISTR02 SUBROUTINES - 5'

49 31/3?/130o FORTRAN (3.1)/MSOS 01/01/71 Si HHOUTTNF UISTRfV NnvEMRFP I^^-PHOGHAMMFM win SlPMOUTlN UISTROÜ OETrgMiNtS MANMMG LEVELS FOR DIsTWlbUTInN AWEAS HASFL) ON luiowlu WIDE AvAlLABILITv OF PEWSONNPL. AI.rUiArFb TOTAL ASSETS UHMENTIY SERVIL IM AN A^EA. ALUtATF.S TOTAL AVAUAHLE FOH PFASSIGNMLNI TO SPEIFI 01STRI HMT IOI^J AWFAS. FOLLOWI NG UTIVE ANI) INTEGER INTEGER INTEGER INTEGER INTEGER IMTFGFR IMTEGEH RFAL IP ITRTIME THINT(100 TROUTSdO f-stulod HSTOLI (1 NFEOS(30) intouh(50 OUTSUb(b() HFP(bO). los(lu) NTODW MOW LSTHSTSd luistun hfgrokdo ENOOLIO TYPE(lOO) NPRLEV NWN0ÜE ( 1 inteb 5TATF AUTH(lOÜ) MATWUGHM DGtJASSEI ( PERIAV( 10 HLL(IOO) LIMITED UGRAUThd ITRAN OVER UTlVEd MRATIVt( TYPE IOS AT HFGOL FSTROKO AUTH FIRST MENUbED,lNEUfl)»IFNU.NOTRdO).LTNGTHd ).TRlNSdOO) 0),F9THÜ»»0( 00).LST0L0( 00)»PTOUTd»NoOfiei NSTT(3O). >«) ion(so)t,neu(j0),np.lsthstodo) 0),ISIJM n) 0).SUR(100).NT no).nott INSllH(bO) AFTFM(bO) IGPAUE(SO.NFFdOO) «tos «ITT»GoPSUMd.ENOROKd»MftTSUMd.NMM(IOO).IHOLO.MTN(IOO).FIRST MFNTS USED ÜN Y IN A.PRAUTHdO.lO) no) «RdOO) 100)»PFRNOSd 0)»MATWRER.MANLEVdOO) losogrdoo) 00)»MANDGRI SFTION 1 ENTRY INPMTU INPUTS UISTHIbUTION DATA ENTRY INPIITD.JK,SYST(AR.4fl) Oo).OUTdOO).TROliTT(lOO) inü).lbthowodoo) 100).FSTHOwKlOO) 00).PTIN(IOO) TNÜSdO.lO) jouttouw(e;0),peh(e,0). PEPOUT(SO) ), ITYPE(«iO).IFIlL.MAXLFM.SYS(4B.4fl).LSTBSfTrlü) MAXSUR.AT(IO) 00) 00).BtGOLdOO) 00).MATGRPSdOO).ATilALdOO),LEN.LEVEL.M.OHPTNPRdoO).LAST.IHRINT SSET.PRFPäRE. AND ALLOT.ASSPTSdOO),PHASSET(10.10) 00).PERNSTdOO) (100).NtWTNdnO).NEWAllTHdOO) 00).NAVAT I AM IN WITH -X 5QA VARIABLE III«no).iTRANdO 0) 10.10),SIJH»SYST.FIRST.TRINT.FSTOLO.PhAUTH.DGRAUTH.MANLFV.ATUAL.OiiTSUb «GRPSUM.TBINS.FSTROXI.ASSETS.SiiMAUTM ENDrOL.OUTTOUR.HEGROW.THOMTT.FSTOLI.F1LI.SYS.AFTER.ENOROw.TROUTS.DGRASSFT RFAD IGO.'oOISGR.NpERAT.NoPKGRp.NMATPlX.LlMlTEn.NAwAT.lPERloD NODISGRxNnMbER OF UISTRIHUTION GROUPS«NPERATsNMMBFH NOPPGRPsNiiMBER NMATRlXrNnMHFR OF PERSONNEL ATEGORTFS. OF PRIORITY GROUPS. OF MATRIes SlMULATEn, «I Preeding page blank loff«h JiN 6 JAN b JAN loff«loffb 71 PAFFBTI

50 NAVATSNUMHER OF AVAILABILITY ATFÜOHIES. PRINT 200. NOUISGR.NPFRAT.NOPMGRP.NMATHIX.MAVAT PARAMFTFR«; TO DETERMlMK AvAlLABlLITy PFRENT Or AVAILABLE FROM ITH MATRIx «HO ARF ELIGIBI E FOR ONUS ONLY = PEHNOS(I). (NOT OVERSEAS) ONUS />NU LT=PEMNST(I). (NOT SHnRT TOUR) GENERAi DISTRIBUTION = PERIAV(I). RFAD lolf (PERNOS(I)»PEuNST(n»PERIAv(I).Isl,NMATRIX) PRINT 20l,(I,PERN0S(I),MERNSr(I) t PERlAV(I),I=i,NMATDlX) if(ios.lr.o) 60 TO 5 InSDGR(I) s OISTRIBUTION GROUP FOR ITH INPUT TATERORY. RFAD 100»(IÜSDGR(I)»I=1,IOS) PRINT 2nB.(l,IOS06R(I),I=l,lOS) PARAHETEW«: TO DETERMINE ASSETS MATWDGR(I»=ÜISTRIBUTI0N GROUP FOR ITH MATRIX. b READ 100» (MATWDGRd) ti = l.nmatrix) MATWPERd)»PERSONNEL ATEGORY FOR T TH MATRIX. RFAD 100.(MATWPER(n»I = l,nmatrix) PRINT 202,(I.MATWDGR(I).MATWPER(I),1=1.NMATHIX) PARAMETER«; TO DETERMINE AUTHORIZATIONS AND MANNING LEVELS R(I)=FILL RATt FOR ITH DlSTRlRUTlON-pRlORITv GROUPIMG. N=N0PR6RP»N0DISGR RFAD 101»(R(I),I=1,N) NnAUTH=NOnISGR«NPERAT»NÜPRÜRP RFAD 100»(AUTH(I),Isl,NOAUTH) RFAD 100.(NEWAUTH(I)»I=FlRSTtLAST) NFWAUTH VFTOR ONTROLS INPUT OF NEw AUTHORIZATIONS. IF NEwAUTH(I) =0, PREVIOUS AUTHORIZATOONS ARF USED. =1, NN AUTHORIZATIONS ARE REAn. M=L = 0 PRINT Isl.NOOlSGR DO 10 J=1,NPEHAT HsO DO 10 Ksl.NOPHGRP L=LM M=M*1 10 PRINT 20*.I»J«K«AUTH(L).R(M) RETURN SETION 2 ENTRY AVAIL ALULATES THE NUMBER OF PERSONNEL AVAIi ABLE FOR REASSIGN. MENT TO THE FOLLOWING DISTRIBUTION GROUPlNGS- NOssONUS ONLY (NOT OVERSEAS) NST«ONUS AND LT (NOT SHORT TOUR) IAv«ONUS.LT, AND ST(GENERAL DISTRIBUTION). ENTRY AVAIL NOS«NSTalAV«0 IF(IOS,LF.O) 60 TO DO 12 I=1.I0S IF <I0SD6R(I).EQ.1) NOS=NOS*IDS(I) IF(I0SD6R(I).EQ.?) NST=NST*10S(I) IF(I0S06R(I).EQ.3) IAV=IAv IDS (I) rniut tin ir 12 ONTINUE 11 DO?0 UI.NMATRIX IUSE«? 4?

51 TMTOT=n ALL SUMM^wr (i?.i.lntot.iostto) NnSsNOS^U'rOT^PtWNOSd) NST=NST»lnTUT»PFHNST(I) IAV=IAV*IMTüT«PEHIAV(I» lf(lmtot«ptwnos(i)*int()t«pehnst(n*tntot»pehlav(i). T.INTOT) IAV = llavmintot-pfhnos(i)«ifnot-pernst(i)«intot-pehlavm»lntot)?0 ONUNUF PPINT 20^. NOS,NSr«IAV PfTURN SFTIÜN J ENTRY AljTwOH GROUPS AUTHORIZATIONS TNPUT HIT OlSTRlRn T ION GROUP, PERSONNEL ATEGOHYS.ANÜ PRIORITY GROUP«AODUINR TO OISTRIRUTIOM GROUP AND PRIORITY GROUP. IT USES SUBROUTINE PRtPAHE. PRAUTHd, i)=authorlzations F 0 R ITH pplority GHQUP WHIH AN BE FILLFO HY THF JTH ÜISTRIHIITION GROUP PFRSONNEL, RMTWY AIITMOR IF inltiai.lf. I, NErt AUTHORIZATIONS ARE READ fnr THIS TIME PERIOD. DTHFRwlSFtAUTHORI/ATlONS ARE THF SAME AS THf PRFVlOUS TIME PERIoD. INITIAL =-> ALL PREP A Rt(INITI AL.NPER A T,NOPRGRP,NOD ISGR) RFTURN SETION h ENTRY ASSrT SUMMARIZES ASStTS FROM MATRIES AORDING TO DISTRIBUTION 6R0iJPs»0LLAPSlN6 OVER PERSONNti ATEGORlLS, ENTRY ASSFT ASSETS(I)=pERSONNEL UPWENTLY SERVING IN ITH MATRIX. DO 40 Isl.NODlSGk nr,rasset(t)so DO *0 Jsl.NOPHGRP 40 PRASSET(J.I)r0 PRASSFTs(I,J)sASS>tTS FOR JTH PRIORITIES WlTHlN ITH DISTRIBUTION GROUPS. DO 45 Ul.NMATRlX JsMATMDGRrI) DGRASSET ( I)sDGRASSET(J)»ATUAL(I) 45 ASSETSmrATUALfU PRINT 207,(J.DGRASSET(j),j=l,NODlSGi) RFTURN SFTION b ENTRY UISTR DlSTRIdUTES ASSETS AMONr, PRIORITY GROUPS WlTHlN EAH DISTRIBUTION GROUP. ENTRY OISTR DO 50 UI.NUDISGR NFWlP«D6R*SSEr(I) IF(NEMIP. E.O) 60 TO 50 LIMITED«! jpr»i»nophghp inis.i NODIS«! NE*IP»PEPS0NNEL TO BF DISTRIBUTED

52 MOOISSNUMHEH OF DISTRIRUTIUN GROUP«; TO HE TONSIDERFD. inis=flht;t nistbihurinn GWOUP TO HF ONSIÜFHtO, loverso nislrihlltr AVAILABLE PERSONNEL UP To AUTHOHIZATION levels. ALL Al.LOT(NEWIP,IUIS,NOPHGHP,NODIS.IOVER) SÜ ONTINUE 00 b5 I=1.NAVAT ifd.le.l» NEWIP=N0S IF(I*0«2) NEWIPsNST IF(I.GE.3I NEWIP=IAV IFINEWIP.I E.o) GO TÜ 55 IOVER=0 IDlSsNAVaT-I*! Nnnis=l L1MITEÜ=1 ALL ALLOT(NEK IP»IDIS.NOPRGHP.NOOIS.IOVER) 55 ONTINUE lfdover.i E.O) GO TO 58 DTSTRIRUTF EXTRA PtusoNNEL TO ONUS. inls=nonl«;gr Moni5=i NFWlP=IOVrW lover=0 LIMlfEDsO ALL ALLOT(NEW IP. IDIS.NOPRGHP.NODIS.IOVER) 58 ONTINUE RETURN SETION f> ENTRY GMMANLEV TOTALS MANNING LEVELS AROSS PRIORITY GROUPS WITH nistuihlltion GROUPS. THE TOTAi S ARE SFf EoUAL TO THE VARIABLE NFEnS, ENTRY GMMANLEV K = 0 DO ft5 I=l.NUnlSGR Or,RMAN=0.n nr,rman»ol<;thibution GROUP MANNING LFVEL, DO ft4 Jal.NÜPHGRP KrK*l 64 0GRHAN«D6RNAN*HANLEV(K) 65 NFn(I)=NEFOS(l)=OGHMAN PRINT 209.(I,NEEI)S(I),I«l.NOüISG»» DO 66 Isl.NMATRIX JsHATWDGR(I) NFEDS( J) IEEDS (J)-ASSETS (I) IF(NEEDS( D.LT.O) NEEDS(J)30 66 ONTINUE PRINT 2lO.(I,NEEOS(I).I»l,NOUIS6H» RETURN SFTION 1 ENTRY OATFTSTOH ALULATES AND STORES TOTALS OF ASSETS AND TRANSIENTS BY DISTRIBUTION GROUP AND PERSONNEL ATEGORIES. ENTRY DATASTOH LsNODISGR«NpEHAT N*NOOISGR«NPERAT DO 7* I«l.N 44

53 74 ITRAN(I)=TATl\/f:(I)=Ü L) 0 7S Isl.NMAfRlX»FAD ITH MATHIX ANU SUM ASSETS, ltrans=iartual=irtaü=n USf = l ALL SUMMARY(1,I,IATUAL»IUSE,IRFAO) II SE = 1$IMFAU=1 Sl'M TRANSIENTS IN HOW 1 ANO OLUMNS l TO MAxLEN OF THE 1TM MATRIX. ALL SUMMARY(1,1,1 TRAMS,lUSE.IREAD) TOTAL ASSFTS ANO TRANSIENTS BY DISTRIBUTION GROljP ANÜ PERSONNEL ATFGOHY. JsMATWDGR ri)«nperat-npehat*matwpeq(i) I AT I VE(J< = I AT I VE{J)*IATUAL 7 ITRAN(J)=TTRAN(J)*ITRANS Mn R E TOTALS OF ASSETS ANO TRANSIENTS QN LU U. K=INTF(3«2-1 ALL RANWRITF(U,IAT1VF»L«K) ALL RANWWITF(14,ITRAN,L»K) PRINT 211.K»(I,IATIVF(1),ITRAN(I)»T=1»L) RFTURN SETION 8 FMTRY OIJTPUTO REAUS STORED DISTRIBUTION DATA ANO OUTPUTS IT IN A MOOIFIFD TARUi AR FORM. ENTRY OUTPUTD f. ipfpsl L=NO0ISGH«NHEHAT DO 80 1=1.L ASSETS(1)=ATUAL(I)=0 BO IATIVE(I)=ITRAN(I)«0 PRINT 212 DO VO IsFlRST,LAST K=I»2-1 ALL RANRFAD ( U, IATI VE.L»K) KsK»l ALL RANRFAD(U,ITRAN,L,K) OLLET ASSETSdATlVE) AND THANSIENTS ITRAN«WITHIN DISTRlRUTION- PERSONNEL ATEGORIES INTO SPEIFIED OUTPllT TlMF INTERVALS(IPERTOn), DO 81 J=',L ASSETS(J)aASSETS(J)»IATIVE(J) 81 ATUAL(J)=ATUAL(J)*ITRAN(J) PRINT 217,(J,ASSETS(J),ATUAL(J) J«l,L) IOUTPERSI/IPEHIOD X1«I XIPERIOOMPERIOD XOUTPERsXl/XIPERIOU PRINT 21R,I0UTPER»X0UTPER ifdoutpeo.ne.xoutper) 60 TO 90 SPREAD ASSETS(ASSETS) AND TRANSIENTS(ATUAL) AMONG PRIORITY GROUPS ANO OUPTUT FOR IPER TIME BLOK, PRINT 213,I^FH KAT.O DO 88 Jsl.NUDlSGR PRINT 21*.J PRINT 21'5 #5

54 00 88 K=1,NPE«AT ITERsO «2 ifdter.tip.l) GO TO 83 IATSIAT»! NEWIP=ASSFTS(IAT) 83 InlSsJ NODIS»! LIMlTEDsO IOVER=0 DO 8* KK=i»NOPHGRP 84 PRASSET(KK»J)=0 IF(NEWIP. E.O) GO TO 91 ALL ALLOT(NEWIP»lOIS.NOPRGRP.NODIS.lOVER) NF(10»10) IS USED TO TFMPROAHILY STORE DISTRIBUTED ASSETS WHILE TRAwSltNTS ARE REING DISTRIBUTED. 91 IF(ITER.GF.I) GO TO KK=1.N0PRGRP 85 PRATIVE(KK»J)SPRASSET(KK,J) ITER=ITER*1 NFWIP=ATMAL(IAT) GO TO 8? IÜM=1,N0PRöHP XPRAUTHsPRAUTM(IOM,J) PFRFILL«POATIVE(IOM,J)/XPRAUTH ISH0RT=XPRAUTH-PRATIVF(I0M,J) 87 PRINT 2l6.K.I0M.PHAUTH(IOM,J),PRArTlVE(IOM.J)»PEBFILL,ISHORT.PR 1ASSET(I0M,J) 88 ONTINUE IPER=IPER*1 DO 89 J>1.L ASSETS(J>3ATUAL(J)«0 89 IATIVE(J)«ITRAN(J)=0 90 ONTINUE RETURN Sl'HRoUTINF DISTRQZ FQRMATS 100 FORMAT(8llO) 101 FORMAT(8Fl0.3) 200 FORMAT(P^MOOISTRIBUTION PARAMETERS//31H0NO. OF OISTBIBUHON GROUPS 1» 13/31M0NO. OF PERSONNEL ATEGORIES «I3/31H0NO. OF PRIORITY GR 101 PS» I3/31H0NO, OF MATRIES MMULATEn > 13 /31H0NO, OF AV IARARILITY AT. «13) 201 FORMATU3M0MATRIX PERENT OF AVAli ABLE ELI6IB E FOR/lOXtl^MONUS 1 ONLY OwUS.LT ONUS.LT,ST/13X«5M»NOS) 5X,5H(NST), 7X.SM(TAV)// l(i5.f12.3,f11.3*f12.3n 202 FORMAT(M7H0MATRIX DISTRIBUTION GROUP PERSONNEL ATEGORY/( ItttlM 203 FORMAT(/SAMOOISTRIBUTION PERSONNEL PRIORITY AUTHORIZATION FIL 1L/36H r,houp ATEGORY GROUP 15X.6HRATI0 /) 204 F0RMAT(I7,I13«II1«IU«F12.3) 205 FORMAT(//«0H0PERSONNEL AVAILABLE FOR DISTRIBUTION TO/21H0ONUS 1 I5/21H0ONUS AND LT «. I5/21Hfl0NilS«LT«AND ST 1=.15) 206 FORMAT<18HOASSETS FOR MATRIX.I4.3H, 15/(18X,I4,3H >.15)) 207 FORMAT(28H0DISTRIBUTION GROUP ASSFTS/(I14.8X,T6)) 208 FORMAT (/2OH0INPUT GROUP OISTR. GROUP/(17» lox» I«il ) 209 FORMAT(12MODISTR. örp.,3x,13hmanninr, LEVEL/(I8»T 16)» 210 FORMAT(12MODISTR. GRP.. 3X,12MNEEDED INPUT/(18.116)) 211 FORMAT nohouata STORED FOR DISTRO PrRlOD«I3/26HoGROi P ATIvE TRA lnsients/(t5»i9 f I12))

55 2\2 F0RMAT(//?7X.25H0lSTPI H UT10N 0*1* SUMMARY//» 213 F09MAT( 3iX»10MTIME BLOK.13) 2U F0RM4T( /19M00ISTRIBUTION 6R0UP»I3) 215 FORMAT (/ QMOPERSONNEL ÜMMANO NUMBER MANNING r'ehent/tzho 1ATEG0RY ELEMENT AUTHORIZED LEVFL FILL SHORTFALL TRÄNSI 1ENTS) 216 FORMAT(I7.I10«IH«F11.1«F10.J«I10«F11.3) 217 FORMAT(22H0MATRIX ATIVE ITRAN/(U.IV,17)) 21«FORMATdOwOlOUTPER»«Il.lOHXOUTPER a F3.2) EMO FORTRAN DIAGNOSTI RESULTS FOR 0ISTRO2 VARIABLES KMIH APPEAP ONLY IN A TYPE STATEMENT SUMAUTH MFNIISED NULL STATEMENT NUMBER«!

56 3l/32/330n FORTRAN (3.1)/MSOS Sl^KOUTINF ALLOT (NEWlFMUIS.NOPHGHP.NOUlStlnVtR) 3 NOVEMbFR SUBROUTIME OMMflN FOLLOlrflNb INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER RFAL 1^69. PROGRAMMER KITT ALLOT DISTRIBUTES ASSETS TO PR IR,INEU8) «IFNO ITRTIME,NOTR(10)»LFNüTMd TRINT(100)tTRlNS(100> TROUTS(lOO),F«;TRÜW0( KSTOVOdOO)»LST0L0( KSTOLI(IOO).PrTOUTd NEEDS(30)<NE(30*10)«NSlT (30)«lNT0UR(5n), IN«;ja(50) OIJTSUB(50)* AFTEH(50) KEP(SO)«100(50)«I6RAOE(50 01/01/71»OUTTOUR(sO) «PEH(SO)«PE«OUT(50) ), ITYPE(SO) los(io) *NEO(30) «NrEdOO) «IFILL «MAXLEN NTOUR»NP ttos,sys(48«48> POW LSTRSTO(IO),LSTPSTT(lO) LSTRSTSdO)»ITT «MAXSUB»AT(30) iniston»isum»grpsumd 00) HEGROWdOO)»ENDROwd 00)»BEGOLdOO) ENOOLdOO).MATSUMd TYPE(lOO).SURdOO) «NiiMdOO) NPRLEV «NT «IHOLÜ NRNOOEdOO)»MTN(IOO) INTE6 «NOTT «FIRST STATEMENTS USED 0N Y IN ft AUTH(lOO) «PRAUTHdO*ln) MATMDGRdOO) «HdOO) DGRASSETdOO) *PFRN0S(I PERIAVdOO) «MATWPEM EILLdOO).MANLEVdOO) LIMITED «losogrdoo) lority GROUPS.,JK «SYST(48t48) 00).OUTdOO) «IROllTTdOO) 100).LST^OwOdOO) InO).FSTROwIdOO) 00).PTIN(IOO) INOSdO.lO) ÜGRAUTHdOO).MftNDGRd 00) «NAVAT RAT(IOO).SuMAURdOO) «SUMASSdOn) TYPE,SU8»ATUAL.ENDOL SYS IOS,SYST «OIITSUH «OUTTOUR «AKTER AT «FIRST «GPPSUM.BEGROW «ENDHOw HEGOL «TRINT»TRINS «THOllTT «TROUTS I-STROHO.FSTOLO»FsTROrtI «FSTrOLI AUTH «PRAUTH ASSETS OüRASSET FIRST «ÜGRAUTH»SilMAUTH «FILL MENUSED «MANLEV N=NODlS»Nr«PKGNP ISTAPTSIOTS^NÜPRRRP-NOPRGRP»! ISTOPSISTAWT^N-I PRINT 10/i.N if(limiteri.eo.l) PRINT 103 N=TOTAL MUMfltR OF phioplty &UlHd)=oUTHüRIZATlONS FOR ASSETS(h=AnjAL URRENTLY P(I)=GRÜiiP I RATIO GROUPS TO WHIH ASSETS PRIORITY GROUP T. SERVING IN GROUP I. InFEÖ 6 JAN 6 JAN 6 JAN InFEB InFEB 71 00) «MATGRPSdOO) «ATliALdOO).LEN «LEVEL»M.GMPTN^RdoO) «LAST «IPRINT ZmFpBTl SSET«PHEPARE«AND ALLOT «ASSFTSdOO).PRASSET(lü«lO) 00) *PERNSTdOO) (100) «NEWTNdoO)..NEWAUTH(IOO) «RE UISTRlBUTED, IIMITEO. IF=1«FILL GROUPS USING AUTHORIZATIONS AS UPPER LIMIT. IF=^.F1LL GROUPS WITH ALL AVAIiAHLE PEOSONMEL, MEWIPsNEw AVAILABLE INPUT. nrtaln SuMMAHY STATISTIS

57 SUMAUTHslTEHrü PRINT 100.<UPRAUTH(I),MHASSET(I)»UtvMI Lr)iSsiDis»Nonis-i IrTSTART-' 00 2 J«T0TS»LDIS TnTASS*TOTAUR=0 ALULATt SUMS OF ASSETS ANÜ AUTH0HTZAT10NS«RATT0S FOR JTM OISTRIBUTION GR RAT(J)=0.n Sl'MAUTHsSMMAUH(J)aSUMASS(J)«0 OH 1 KsltuOPRGRP FILKD.O Sl. MAUTH«Si MAUTH*PRAUTH(K»JJ SUMAURtJJsSUMAURJl^PRAUTHKfJj^Rd) 1 Sl'MASS(J)=SUMASS(J)*PPASSET(K,J) PRINT 106, SUMAUTH»SUHAUR<J)»SUMASS(J)»TOfASS IF(SUMASS(J),LE.0) PRINT lol.rat(j) ALULATE TOTAL NUMBER OF ASSETS ANf) AUTH«RATIOS FOR ENTIRE SYSTEM. TOTASS«T0TASS»SUMASS(J) 2 TOTAUR«T0TAUR*SUMAUH(J) UO PRINT 106,I»TÜTAUR»TOTASS,NEI((IP TOT»TOTASs*NEWlP$TOTAaTOTAUH IF(TOT.LE.O) ÜO TO 41 3 RATIOsTOT/TOTA PRINT 108,RATIO IRUN.I DO J»IOTS»LÜIS IF(RAT(J).NE,0,0) GO TO IF(SUMASS(J).GT.O.O) GO TO 7 RX»0,0 GO TO 8 7 RX«SUMASS(J)/SUMAUR(J) 8 IF(RX.LT.PATIÜ) GO TO 4 WHEN RAT(.I).GT.RATI0«SET PAT (J> «SUM&SS (J)/SuMAUR (J). RAT(J)«RX TOT «TOT-SUMASS(J) TnTA«TOTA-SUMAUR(J) IRUN.2 4 ONTINUE GO T0(5,3MIRUN 5 DO 6 JBI0TS«LÜIS WHEN RAT( )>.LT.RATIO.SET RAT(J)»RATIO. IF(RAT(J).EO.0.0) RAT(J)»RATIO 6 ONTINUE PRINT lll.ufratu) tj>tols«luis) MINIMUM ASSIGNED TO GROUP(I)«ASSETS. IF LlHlTED=i»MAXIMUM ASSIGNED TO 6ROUP(I)«AUTH(1), OTHERWISF.UPPER LIMIT«F(AVAILABLF PERSONNEL), PRINT 108,RATIO MENUSED>0 ITER«ITER»1 UISTART-1 DO 15 J>InIS,LDIS DO 15 K>1,N0PR6RP I«IM IF(FILL(I».E0.1) 60 TO

58 ArPkAUTHUf J) MANLEV(T»r«AT(J)«H(I)«A PPlNT lo'v.i'manlevd) IF(MANLFVM).LT.PHASSFKK,J)) 60 TO 12 if(manlevrl).lt,prauth(kfj)) GO TO 14 if(limiten.eo.l) MANLFV(I)»PRAUTM(K,J) PPINT 109.I,MANLEV(I) IF(MANLEV(I).LT,PRAS5ET(K,J)) 60 TO 12 FTLUDsl U 12 MANLEV(I)sPHASSET(K»J) FTLKD«! 1* MFNUSED«MFNUStO*MANLEV(I) 15 ONTINUE PPINT l07.iteh,menusen PRINT 110.TOT HEK FOR HOUNDING ERRORS IF(TOT.EO.MENUSEO) GO TO 30 IF(TOT.LE.MENUSEO) 60 TO 17 PRINT 110,TOT PRINT ln7,iteh,menuse0 17 NFlLLsO lf(limlten,eq.o) GO TO 2* PRINT 107,ITEM,MENUSED I.ISTART J«InIS,LDIS DO 20 K«l,NOPHGRP NFILL«NFII L»FILL(I) lf(fill(iwne.l) GO TO 20 AsPRAUTH(K*J) TOTA«TOTA-A*R(I) TOT»TOT-MaNLEV(I) 20 ONTINUE IF(T0T.LE.O) GO TO 30 IFINFILL.^T.O^ND.NFILULT.N) GO TO 1 lf(nfill,ro,n,and*tot.eq.menuseo) 60 TO 30 2* TOT «TOT -MENUSEO ITOT.TOT DO 25 I«1,IT0T 25 MANLEV(n=MANLEV(I)*l 30 I0VER«T0T -MENUSEO PRINT 10* I.ISTART-1 DO 40 J«InIS,L0IS DO 40 K»1,N0PN6RP I»I*1 NFWlN(I)»MANLEV(I)-PHASSET(K,J) IF(NEWIN(T),LT.0) NEWIN(I)»0 IF(MANLEV(I).6T.PRASSET(K,J)) PRASSFT(K,J)«MANLFV(I) 40 PRINT 105,(I,R(I),PRAUTH(K,J),MANLEv(I)«PRASSET(K,J)«NEMJN(T)) 41 ONTINUE SUBROUTINE ALLOT FORMATS. 100 FORMAT(2liO«F10,3) 101 FORMAT(8Fi0.4) 50 -