path: root/benchmarks/dcn19.dat

BEGIN NEW DATA CASE
C     BENCHMARK DCNEW-19
C    18 February 1997,  two new features enter the UTPF.  These will be verified
C    by the following data.   See April newsletter for details about:  1) Walter
C    Powell's analytical type-10 source;   2) $PARAMETER  for both simple string
C    substitution and also mathematics at data-input time (Walter Powell pocket
C    calculator).  The Type-10 function can be put in cols. 11-60 if it fits, or
C    can be continued onto cols. 11-40 of the following card ordered by CONT. in
C    columns 61-61 of the first.  T-start and T-stop are always from ending card
C       August of 1998: replace Walter Powell's pocket calculator by one written
C       in Vancouver, USA.  Story is scheduled for the October newsletter.
$PARAMETER START ONLY   { This block of 3 cards is used only before any $INCLUDE
FIVEK= '5.E+3' { Within apostrophe ==> there is no math, just string replacement
BLANK card ends  $PARAMETER  definitions
     1.0     21.
       1       1
$PARAMETER   { This will be serviced by CIMAGE just as any other $-card would be
C ONEOHM = 0.5 * ( 1.0 + "DELTAT" ) { Within quotes ==> value of this ATP variable
ONEOHM = 0.5 * ( 1.0 + DELTAT ) { Beginning 28 Nov 98, quotes are no longer needed
C OMEGA = "TWOPI" * .05  ---  original frequency,  equal to that of Type-14 COS
C OMEGA_ = "TWOPI" * 0.1   { Speed the modulation of the ramp for visual effect
OMEGA_ = TWOPI * 0.1  { Beginning 28 November 98, quotes no longer are necessary
C ANGLE_ = .25 * "TWOPI"
ANGLE_ = .25 * TWOPI  { Beginning 28 November 98, quotes no longer are necessary
BLANK card ends  $PARAMETER  definitions that are processed just b4 branch cards
  COS                     ONEOHM                                               1
  SIN                     ONEOHM                                               1
  SIN2                       1.0
  SIN3                       1.0
  A1    A2                 5.E+3                                               1
  A2                       FIVEK
BLANK card ends branch cards
BLANK card ends switch cards (none, for this case)
14COS            1.0       .05       0.0
C 14SIN            3.0       .05      -90.     ----  old form of following card:
C 10SIN     3.0 * COS ( OMEGA * TIMEX - ANGLE )  { Walter Powell analytical source
10SIN     .1 * TIMEX + .8 * COS ( OMEGA_ * TIMEX - ANGLE_ )  { Walter Powell analytical source
10SIN2    .1 * TIMEX + .8 * COS ( OMEGA_ * TIMEX - ANGLE_ ) CONT.   { Note 61-65
          + 0.2 { continuation carries T-start, T-stop in 61-80 }  1.5      18.5
C    Note the preceding  SIN2  repeats SIN but with extra vertical offset of 0.2
C    on a continuation card.   Also,  SIN2 uses T-start (61-70) & T-stop (71-80)
10SIN3    IF( TIMEX .LE. 15. ) THEN     { If simulation time < 16 sec:
          SIN3 = TIMEX * 0.10           { Linear ramp up with slope matching SIN
          ELSE                          { Alternatively (if 16 or more seconds):
          SIN3 = ( 23.5 - TIMEX ) * .20 { Linear ramp down,  double the slope up
          ENDIF                       { Terminate 5-line block of Type-10 source
C    Note the preceding  SIN3  creates a saw tooth with the peak broken off to
C    create a flat top.  The value at 16 is equal to the value at 15.  Note that
C    the name of the bus in columns 3-8 of the Type-10 source card is repeated
C    within the IF-block (left of equal signs).  Control words  IF(  and  ELSE
C    and  ENDIF  must begin in column 10 exactly as illustrated.  As added on
C    11 December 2000,  ELSEIF(  also should be legal,  but is not being
C    illustrated in this data.
14A1         620000.       .06      -50.
BLANK card ends source cards
  SIN3  SIN   SIN2  COS   A2
C  First  5     output variables are electric-network voltage differences (upper voltage minus lower voltage);
C  Next   3     output variables are branch currents (flowing from the upper node to the lower node);
C  Step    Time  SIN3       SIN        SIN2       COS        A2         COS        SIN        A1
C                                                                       TERRA      TERRA      A2
C    0     0.0     0.0        0.0        0.0        0.0        0.0        0.0        0.0        0.0
C    1     1.0     0.1 .570084377        0.0 .951056516 272691.058 .951056516 .570084377 54.5382115
C    2     2.0     0.2 .960785678 1.16078568 .809016994 307819.308 .809016994 .960785678 61.5638615
C    3     3.0     0.3 1.06090928 1.26090928 .587785252  299715.25 .587785252 1.06090928 59.9430501
C    4     4.0     0.4 .870407977 1.07040798 .309016994 249517.077 .309016994 .870407977 49.9034154
C    5     5.0     0.5 .500237061 .700237061  .6123E-16 164274.972  .6123E-16 .500237061 32.8549944
C    6     6.0     0.6 .129975602 .329975602 -.30901699  55960.935 -.30901699 .129975602  11.192187
C    7     7.0     0.7 -.06076275 .139237247 -.58778525 -60212.649 -.58778525 -.06076275  -12.04253
C    8     8.0     0.8 .039067835 .239067835 -.80901699 -167929.55 -.80901699 .039067835 -33.585909
C    9     9.0     0.9  .42953207  .62953207 -.95105652 -252061.24 -.95105652  .42953207 -50.412247
BLANK card ends requests for program outputs (here, just node voltages)
C   10     10.     1.0 .999688816 1.19968882        -1. -300791.68        -1. .999688816 -60.158335
C   11     11.     1.1 1.56996442 1.76996442 -.95105652 -307276.82 -.95105652 1.56996442 -61.455363
C   12     12.     1.2 1.96073982 2.16073982 -.80901699 -270605.84 -.80901699 1.96073982 -54.121169
C   13     13.     1.3 2.06095504 2.26095504 -.58778525 -195929.08 -.58778525 2.06095504 -39.185817
C   14     14.     1.4 1.87052788 2.07052788 -.30901699 -93734.666 -.30901699 1.87052788 -18.746933
C   15     15.     1.5 1.50038531 1.70038531 -.1837E-15 21624.5069 -.1837E-15 1.50038531 4.32490137
C   16     16.     1.5 1.13009557 1.33009557 .309016994 133946.582 .309016994 1.13009557 26.7893164
C   17     17.     1.3 .939283118 1.13928312 .587785252 227456.258 .587785252 .939283118 45.4912516
C   18     18.     1.1 1.03902209 1.23902209 .809016994 289020.378 .809016994 1.03902209 57.8040757
C   19     19.     0.9 1.42941218        0.0 .951056516 309992.446 .951056516 1.42941218 61.9984891
C   20     20.     0.7 1.99954057        0.0        1.0 287426.995        1.0 1.99954057  57.485399
C   21     21.     0.5 2.56984444        0.0 .951056516 224493.277 .951056516 2.56984444 44.8986554
  CALCOMP PLOT
 143 2. 0.0 20. -1. 3.0 COS   SIN   SIN2  SIN3
BLANK card ending plot cards
BEGIN NEW DATA CASE
C     BENCHMARK DCNEW-19
C      2nd of 6 subcases is related to preceding.  Analytical function is too
C      long,  spilling over into T-start of columns 61-70.  Demonstrate special
C      handling to reject such data in civilized fashion.   28 October 1998
     1.0     21.
       1       1
C  19 March 2009,  WSM adds 2 unrelated symbol definitions  HANSOK and  HANSBAD.
C  But first,  add  a request to prohibit optimization by the pocket calculator.
C  Addition here is mandatory  until an error in optimization of the compilation
C  is corrected.  The troubled data are  formulas  for  HANSOK  and  HANSBAD  as
C  first reported by  Prof. Hans Kr. Hoidalen in Trondheim, Norway.  Without the
C  change making  NOOPT = 1,  execution would end on the HANSBAD input card with
C  an extraneous complaint about TACS (not being used).  With use of  NOOPT = 1,
C  the computation proceeds normally.  Although this is not  HFS data,  the loop
C  counter should have default value  KNT = 1  so  the right hand side correctly
C  evaluates to  10**1 = 10.   Execution still will end prematurely with that
C  complaint about too long an analytical function for  SIN,  but at least Prof.
C  Hoidalen's data is handled correctly.  WSM.
C  $PARAMETER   NOOPT=1 { Set optimization of pocket calculator compilation to none 
C    Comment out preceding line  1 October 2011 after correction was made to the
C    code that handles  TEMPx  variables within  POCKET.SPL  and nowhere else.  
$PARAMETER   { This will be serviced by CIMAGE just as any other $-card would be
OMEGA_ = "TWOPI" * 0.1   { Speed the modulation of the ramp for visual effect
ANGLE_ = .25 * "TWOPI"
C  Note about preceding quotation marks on  24 December 2000.  Whereas the first 
C  subcase demonstrated removal long ago,  this use is retained in its original, 
C  inefficient form as a demonstration of a new property.  With the introduction 
C  of non-volatile variables,  the quoted  TWOPI  continues to work as expected,
C  but it now demonstrates a new power.  The way MATDAT handle quotes,  they are
C  erased, and the line is preceded by a temporary definition TWOPI = 6.2832 ...  
C  which is taken as a SPY symbol (a variable in COMMON, recognized by SPY).  In
C  fact the answer is correct,  but  COMMON is being modified without the user's
C  knowledge or intention.  Old inefficiency (requiring a intermediate variable)
C  suddenly has been increased by unnecessary redefinition of a COMMON variable.
HANSOK  =10.**((KNT-1.)/20.+1.)     { works.  1st of 2 formulas from HKH
HANSBAD =10.**((KNT-1.)*0.05+1.)    { fails   2nd of 2 formulas from HKH
BLANK card ends  $PARAMETER  definitions that are processed just b4 branch cards
  COS                        1.0                                               1
  SIN                        1.0                                               1
BLANK card ends branch cards
C  2 April 2009,  WSM adds unrelated IF-THEN block that includes an intermediate
C  variable only.  Trouble using IF without a permanent variable was reported by
C  Prof. Hans Kr. Hoidalen in Trondheim, Norway. Although this is not  HFS data,
C  the loop counter should have default value KNT = 1 making the  IF-THEN  block
C  legal.  Add a dummy  switch card  to illustrate use of the permanent variable
C  CAP___ which depends on intermediate variable CAPI.  Execution still will end
C  prematurely with that complaint about too long an analytical function for the
C  variable SIN, but at least Prof. Hoidalen's IF-THEN will be handled properly.
C     Note about the following  IF(.  The use of  "equals one"
C     is arbitrary.  Changing  .EQ.  to  .NE.  will force control
C     through the  ELSE  alternative,  resulting in  CAPI = 5.0  rather
C     than  2.3,  and as a result,  CAP___ = 10  (see next line) :
C       Parameter 1  defined.  Value =  1.000000E+01    |CAP___ = 2.0 * CAPI
$PARAMETER   { This will be serviced by CIMAGE just as any other $-card would be
IF( KNT .EQ. 1.0 ) THEN { This is equivalent (THEN on right is not required)
CAPI = 2.3  $$     { The double dollar sign flags an intermediate variable, note
ELSE     { Alternatively (if not the first pass, so for KNT = 2 onward):
CAPI = 5.   $$     { The double dollar sign flags an intermediate variable, note
ENDIF   { Terminate 5-line block that defines intermediate symbol CAPI
CAP___ = 2.0 * CAPI { Define A6 permanent variable to be used in later data
BLANK card ending $PARAMETER block
C   Note about multiple intermediate variables within the  IF-THEN  block:
C   the logic probably would fail.  As now constituted,  one intermediate
C   variable is the limit within an  IF-THEN  block.  But more than one  IF-THEN
C   block is possible.  To illustrate,  let the desired second intermediate
C   variable  CAPJ  be defined within its own  IF-THEN  block as follows.  Also
C   for variety,  we will change  ".EQ. 1.0"  to  ".NE. 1.0"  to force the use
C   of the  ELSE   alternative: 
$PARAMETER { Separate $PARAM block ==> correct interpret line count within block
C   Added explanation of preceding line.  Without it and preceding BLANK line,
C   the answer would be correct except for line # of interpretation, which would
C   not return to value 1 for the IF( statement. Restarting this at 1 is clearer
C   is clearer for the user who looks closely at such details.  Numbering of the
C   interpretation then will be  1, 3, 4, and 5  instead of  3, 5, 6, and 7.
IF( KNT .NE. 1.0 )
CAPJ = ( KNT + 0.5 ) ** 2 $$     { For KNT equal to 1, this gives CAPJ = 2.25
ELSE     { Alternatively (if not the first pass, so for KNT = 2 onward):
CAPJ = ( KNT - 0.5 ) ** 2 $$     { If KNT were equal to 1, this would give 0.25
ENDIF    { Terminate 7-line block that defines symbols CAPI and CAPJ
IND___ = CAPJ / 2.0        { Inductance in cols. 33-38 of following series R-L-C
BLANK card ending $PARAMETER block
C  The following meaningless switch illustrates use of CAP___ & IND___ variables
  COS   SIN       CAP___    IND___
BLANK card ends switch cards
14COS            4.0       .05       0.0
C                              Ruler for T-start in 61:70:  1234567890
10SIN     0.1 * TIMEX + 0.8 * COS ( OMEGA_ * TIMEX - ANGLE_ ) { Too long function
BLANK card ends source cards
  SIN   SIN2  COS   A2
BLANK card ends requests for program outputs (here, just node voltages)
BLANK card ending plot cards
BEGIN NEW DATA CASE
C    3rd of 6 subcases illustrates Type-10 sources only.  dT-loop output will be
C    identical to that produced by famous TACS and MODELS simulations  (learning
C    how slow code from Laurent Dube really is).  With extension to 100K steps
C    and proper control of output,  this becomes  MATHTY10  as first used to
C    time operation of  POCKET  (WSM's new pocket calculator).
C      That might have been 1998. Twelve years later, the inefficiency of MODELS
C      is widely known. So distorting the data a little is permissible. Addition
C      of the GAUSS function to both TACS and the pocket calculator will require
C      verification.  DC-18 does the work for TACS.  Here,  on  10 October 2010,
C      we add such disconnected, unrelated output to the pocket calculator. WSM.
PRINTED NUMBER WIDTH, 10, 2,  { Request maximum precision (for 8 output columns)
     .02     2.0 
       1       1       0       0       1      -1
       5       5
  TEST1                      1.0
  TEST2                      1.0
  TEST3                      1.0
  TEST5                      1.0
  TEST6                      1.0
  TWO?                       1.0
  NORMAL                     1.0   { A Gaussian random # will be applied to this
BLANK card ends branch cards
BLANK card ends switch cards (none, for this case)
10TEST1    10.0 * ( 1.0 + TIMEX ) ** 2 + 50. 
10TEST2    1.E2 * COS ( 2.0 * 3.14159 * TEST1 / 100. )
10TEST3    10.0  +  5.5 * TIMEX * SQRT ( ABS ( TEST2 ) )
10TEST5    2.0 * 3.14159 * TEST1 / 100.
10TEST6    2.0 * SIN ( TEST5 )
C 10TWO?     2.E-4 * ( TEST2 ** 2 + 2500. * TEST6 ** 2 )
10TWO?     2.000E-4 * ( TEST2 ** 2 +                        CONT.
           2500. *  { Remember that T-start and T-stop are on right of this card
     1     TEST6 ** 2 ) { Illustrate conventional continuation line (1 in column
C                         6 preceded by 5 blanks).  With  or without CONT.,  any
C                         number of continuation cards can follow the original 1
C                         or two  (two if CONT. is used).  This is like FORTRAN.
10NORMAL   GAUSS ( 0.0 )  { Random function has zero mean and stand. deviation 1
BLANK card ending all sources
  TEST1 TEST2 TEST3 TEST5 TEST6 TWO?  NORMAL
C  Step     Time     TEST1     TEST2     TEST3     TEST5     TEST6     TWO?      NORMAL
C     0      0.0        0.0       0.0       0.0       0.0       0.0       0.0       0.0
C     1      .02     60.404   -79.384  10.98007  3.795292  -1.21625       2.0   -.26382
C     2      .04     60.816  -77.7833  11.94029  3.821179  -1.25694       2.0  .6641848
C     3      .06     61.236  -76.0979  12.87873  3.847568  -1.29755       2.0  -.384642
BLANK card ending node voltage outputs
C    90      1.8      128.4     -21.2  55.58307  8.067603  1.954539       2.0  1.195179
C    95      1.9      134.1  -54.1115  86.87076  8.425744  1.681897       2.0   .146522
C   100      2.0       140.  -80.9013  108.9396  8.796452  1.175583       2.0  .9643848
C Variable maxima :    140.  99.99389  108.9396  8.796452  1.999977       2.0  3.103447
C Times of maxima :     2.0      1.24       2.0       2.0      1.74       .04      1.68
C Variable minima :     0.0  -80.9013       0.0       0.0  -1.99999       0.0  -2.77845
C Times of minima :     0.0       2.0       0.0       0.0       .58       0.0       .82
  CALCOMP PLOT  
 143 .2 0.0 2.0         TEST1 TEST2 TEST3
 143 .2 0.0 2.0         TEST5 TEST6 TWO?
BLANK card ending plot cards
BEGIN NEW DATA CASE
C    4th of 6 subcases documents pocket calculator applied to many different
C    expressions.  Use Type-10 sources,  and take just 1 time step.  Prior to
C    16 November 2000,  this was separate disk file  POCKET.DAT     Following
C    diagnostic line should be activated to see all assembler (precedes "EOF").
C    For philosophy of this data,  see the April, 2001, newsletter.
C DIAGNOSTIC           0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 9 9 9 9 9
C    The following illustrates the ability to change dimensions (offset
C    subscripts for VOLTI storage) of the pocket calculator.  Prior to
C    12 November 2000,  dimensions were 10 and 20,  so the following example
C    shows what resulted: an error stop do to inadequate INDOFF.  That was
C    prior to expansion of the default values of 60 and 120,  which handle
C    the present data subcase without difficulty.
C $PARAMETER   INDOFF=10  KONST=20   { Fixed limits prior to improvement 12 Nov 00
C Halt in  POCKET.  Overflow JTOKEN into INDOFF.   INDOFF, KCONST =   10   36
C  Temporary error stop in  ENTRY STOPTP of "WINDOW".     NCHAIN, LASTOV =      5     4
C $PARAMETER   INDOFF=60  KONST=120  { Do-nothing reassignment of default values
$PARAMETER   INDOFF=90  KONST=180  { Do-nothing reassignment of default values
C      Overhaul of storage for pocket calculator requires more List 4 space than
C      previously.  Add  NEW LIST SIZES  on 6 August 2010.  List 4 = 140 fails :
NEW LIST SIZES      { Request for a change to program table sizes (VARDIM input)
DEFAULT      3.0    { Any blank field will be converted to 3 times default value
       0       0       0     240      { List 4 < 840 of LISTSIZE.BPA is adequate
BLANK card for lists 11-20.
BLANK card for lists 21-30.
     390     300     { Optional card for Lists 31, etc. --- as LISTSIZE.BPA
  240000     742     { Offsets for supporting programs (non-simulation overlays)
PRINTED NUMBER WIDTH, 11, 1,  { Request maximum precision (for 8 output columns)
    .001    .001
       1     -11       0       0
C      Each analytical function will define a Type-10 source.  In turn, this
C      must be connected to a node of the electric network.  So, for each,
C      connect a 1-ohm resistor to ground:
C   1st set of 10 variables ANG* is for 1st row of 132-column output vector:
  ANG1                       1.0
  ANG2                       1.0
  ANG3                       1.0
  ANG4                       1.0
  ANG5                       1.0
  ANG6                       1.0
  ANG7                       1.0
  ANG8                       1.0
  ANG9                       1.0
  ANG10                      1.0
C   Next set of 10 variables BAL* is for 2nd row of 132-column output vector:
  CC                         1.0
  DD                         1.0
  THETA                      1.0
  OMEGA                      1.0
  BAL1                       1.0
  BAL2                       1.0
  BAL3                       1.0
  BAL4                       1.0
  BAL5                       1.0
  BAL6                       1.0
  BAL7                       1.0
  BAL8                       1.0
  BAL9                       1.0
  BAL10                      1.0
C   Next set of 10 variables CAT* is for 3rd row of 132-column output vector:
  UNITY                      1.0
  ZERO                       1.0
  CAT1                       1.0
  CAT2                       1.0
  CAT3                       1.0
  CAT4                       1.0
  CAT5                       1.0
  CAT6                       1.0
  CAT7                       1.0
  CAT8                       1.0
  CAT9                       1.0
  CAT10                      1.0
C   Next set of 10 variables DUM* is for 4th row of 132-column output vector:
  DUM1                       1.0
  DUM2                       1.0
  DUM3                       1.0
  DUM4                       1.0
  DUM5                       1.0
  DUM6                       1.0
  DUM7                       1.0
  DUM8                       1.0
  DUM9                       1.0
  DUM10                      1.0
C   Next set of 10 variables EAT* is for 5th row of 132-column output vector:
  EAT1                       1.0
  EAT2                       1.0
  EAT3                       1.0
  EAT4                       1.0
  EAT5                       1.0
  EAT6                       1.0
  EAT7                       1.0
  EAT8                       1.0
  EAT9                       1.0
  EAT10                      1.0
C   Next set of 10 variables FAT* is for 6th row of 132-column output vector:
  FAT1                       1.0
  FAT2                       1.0
  FAT3                       1.0
  FAT4                       1.0
  FAT5                       1.0
  FAT6                       1.0
  FAT7                       1.0
  FAT8                       1.0
  FAT9                       1.0
  FAT10                      1.0
C   Next set of 10 variables GAS* is for 7th row of 132-column output vector:
  GAS1                       1.0
  GAS2                       1.0
  GAS3                       1.0
  GAS4                       1.0
  GAS5                       1.0
  GAS6                       1.0
  GAS7                       1.0
  GAS8                       1.0
  GAS9                       1.0
  GAS10                      1.0
BLANK card ends branch cards
BLANK card ends switch cards (none, for this case)
C   1st set of 10 variables ANG* is for 1st row of 132-column output vector:
10ANG1     1.0
10ANG2     -1.0
10ANG3     ANG1
10ANG4     +ANG1
10ANG5     COS ( 1.0 )
10ANG6     SIN ( ANG3 )
10ANG7     SIN ( -ANG3 )
10ANG8     - SIN ( - ANG3 )
10ANG9     - ANG3 * ANG6 
10ANG10    - SIN ( ANG3 )
C   Next set of 10 variables BAL* is for 2nd row of 132-column output vector:
10CC       3.14159
10DD       2.71828
10THETA    CC
10OMEGA    DD
10BAL1     ( -CC + DD ) / ( -CC - DD )
10BAL2     ( -CC - DD ) / 3.1415
10BAL3     ( -CC + DD ) / ( CC - DD )
10BAL4     -( -CC + DD ) / ( CC - DD )
10BAL5     THETA * OMEGA
10BAL6     THETA + OMEGA
10BAL7     THETA - OMEGA
10BAL8     THETA / OMEGA
10BAL9     THETA + OMEGA + ANG5
10BAL10    THETA + OMEGA + ANG5 - TIMEX
C   Next set of 10 variables CAT* is for 3rd row of 132-column output vector:
10UNITY    1.0
10ZERO     0.0  { In addition to these, remember CC = 3.14159 and DD = 2.71828
10CAT1     CC .EQ. DD       { This is false, so CAT1 = 0.0
10CAT2     CC .NE. DD       { This is true,  so CAT2 = 1.0
10CAT3     CC .GT. DD       { This is true,  so CAT3 = 1.0
10CAT4     CC .LT. DD       { This is false, so CAT4 = 0.0
10CAT5     CC .GE. DD       { This is true,  so CAT5 = 1.0
10CAT6     CC .LE. DD       { This is false, so CAT6 = 0.0
10CAT7     .NOT. UNITY      { UNITY is true, so toggle it to false: CAT7 = 0.0
10CAT8     UNITY .OR. ZERO  { CAT8 = 1 since UNITY is true, & OR with anything is true
10CAT9     UNITY .AND. ZERO { CAT9 = 0 since ZERO is false, & AND with anything is false
10CAT10    UNITY .NOR. ZERO { CAT10 = 0 since OR first, then negate, so toggle CAT8
C   4th set of 10 variables DUM* is for 4th row of 132-column output vector:
10DUM1     UNITY .NAND. ZERO { DUM1 = 1 since AND first, then negate, so toggle CAT9
10DUM2     THETA / OMEGA + DD { DUM2 = 3.87400715 = 3.1416 / 2.718 + 2.718
10DUM3     THETA / OMEGA - DD { DUM3 = -1.562 = 3.1416 / 2.718 - 2.718
10DUM4     CC**DD             { DUM4 = 22.459 = 3.1416 ** 2.718
10DUM5     50.0 + CC**DD      { DUM5 = 72.459 = DUM4 + 50
10DUM6     TRUNC ( CC ) + INVRS ( CC ) { DUM6 = 3.318 = 3 (integer in 3.14) + 1 / 3.14
10DUM7     DEG ( CC )         { DUM7 = 180 (close) = 3.1416 converted to degrees
10DUM8     RAD ( DUM7 )     { DUM8 = 3.1416 = DUM7 deg converted back to radians
10DUM9     SIGN ( -CC ) - 3.0  { DUM9 = -4 = -1 - 3 = sign of ( -3.14 ) - 3
10DUM10    MINUS ( CC )      { DUM10 = -3.1416 = minus sign applied to Pi
C   5th set of 10 variables EAT* is for 5th row of 132-column output vector:
10EAT1     SEQ6 ( DUM4 )  { EAT1 = 4 since DUM4 = 22.4590591 / 6 ==> 4.459 remainder
10EAT2     - LOG ( DD )   { EAT2 = -1 (close) since DD = 2.71828 is close to e
10EAT3     LOG10 ( 200. ) { EAT3 = 2.3010 since log(10) of 2 = .3010
10EAT4     ATAN ( UNITY ) { EAT4 = .785398 = atan ( 1.0 ) = pi / 4 = 3.1416 / 4
10EAT5     EXP ( 2.0 * UNITY ) { EAT5 = 7.389 = 2.71828 ** 2
10EAT6     ASIN ( -ONEHAF )    { EAT6 = -.5236 = -30 degrees / 57.29578
10EAT7     ACOS (  ONEHAF )    { EAT7 = 1.0471 =  60 degrees / 57.29578
10EAT8     RAN ( ZERO )        { EAT8 = .813467631 (arg = 0 ==> next random #)
10EAT9     TANH ( UNITY )      { EAT9 = .761594156 
10EAT10    SINH ( UNITY ) / COSH ( UNITY ) { EAT10 = EAT9 by trig definition
C   6th set of 10 variables FAT* is for 6th row of 132-column output vector:
10FAT1     TAN ( CC / 6.0 )  { FAT1 = .5773 since arg CC / 6 = Pi/6 = 30 degrees
10FAT2     2.0 * COTAN ( 60. / 57.29578 ) { FAT2 = 1.1547 = 2 * FAT1 
10FAT3     1.0 / ABS ( UNITY * CC * DD )  { FAT3 = .1171 = 1 / ( 3.14 * 2.718 )  
10FAT4     - SQRT ( 2.0 ) * 0.5  { FAT4 = -.707 = sqrt ( 2 ) * 1/2  
10FAT5     - 2.0 ** 2 + UNITY    { FAT5 = -3 = -4 + 1  
10FAT6     (-2.0) ** 2 + UNITY   { FAT6 = 5 = 4 + 1  
10FAT7     - EXP ( UNITY * CC / DD ) - 0.1  { FAT7 = -3.276 = -EXP ( 3.14 / 2.718 ) - 0.1
10FAT8     ( UNITY / DD * CC ) * 2.0 - UNITY { FAT8 = 1.311 = ( 3.1416 / 2.718 ) * 2 - 1
10FAT9     UNITY - ( UNITY * DD + CC ) { FAT9 = -4.860 = 1 - ( 3.14 + 2.718 )
10FAT10    UNITY * DD - CC { FAT10 = -.4233 = 2.718 - 3.14 = 1 * 2.718 - 3.1416
C   7th set of 10 variables GAS* is for 7th row of 132-column output vector:
10GAS1     UNITY * ( CC / DD / UNITY ) { GAS1 = 1.156 = 3.14 / 2.718 = Pi / e / 1
10GAS2     CC ** 2 + DD ** 2           { GAS2 = 17.26 = 3.14 ** 2 + 2.718 ** 2
10GAS3     CC ** 2 - DD ** 2           { GAS3 =  2.48 = 3.14 ** 2 - 2.718 ** 2
C              3 December 2001, add variables to confirm correction of the error
C              first reported by Orlando Hevia using data EXP.DAT    Note GAS4
C              through GAS9 all have same value  EXP(-T) = 1 / EXP(.001) = .999
10GAS4     1.0 / EXP ( TIMEX ) { This establishes the right answer (no bug here)
10GAS5     EXP(-(TIMEX*1.0))   { Hevia showed that this, too, always was correct
10GAS6     EXP ( -TIMEX )      { Hevia showed that this, too, always was correct
10GAS7     EXP( -TIMEX * 1.0 )   { Prior to correction on 3 December, error stop
10GAS8     EXP( -TIMEX / 1.0 )   { Prior to correction on 3 December, error stop
10GAS9     EXP( -TIMEX * UNITY ) { Prior to correction on 3 December, error stop
10GAS10    28.0 - 25.0 - ( TIMEX - 1.0 )  { Verify correction on  9 January 2002
BLANK card ending all sources
  ANG1  ANG2  ANG3  ANG4  ANG5  ANG6  ANG7  ANG8  ANG9  ANG10
  BAL1  BAL2  BAL3  BAL4  BAL5  BAL6  BAL7  BAL8  BAL9  BAL10
  CAT1  CAT2  CAT3  CAT4  CAT5  CAT6  CAT7  CAT8  CAT9  CAT10
  DUM1  DUM2  DUM3  DUM4  DUM5  DUM6  DUM7  DUM8  DUM9  DUM10
  EAT1  EAT2  EAT3  EAT4  EAT5  EAT6  EAT7  EAT8  EAT9  EAT10
  FAT1  FAT2  FAT3  FAT4  FAT5  FAT6  FAT7  FAT8  FAT9  FAT10
  GAS1  GAS2  GAS3  GAS4  GAS5  GAS6  GAS7  GAS8  GAS9  GAS10
C Step   Time  ANG1  ANG2  ANG3  ANG4    ANG5       ANG6       ANG7      ANG8        ANG9       ANG10
C   1   .1E-2  1.0   -1.   1.0   1.0 .540302306 .841470985 -.84147098 .841470985 -.84147098 -.84147098
C        BAL1       BAL2    BAL3   BAL4       BAL5    BAL6     BAL7      BAL8      BAL9       BAL10
C    .072238804 -1.8653096   -1.    1.0  8.53972127  5.85987  .42331 1.15572715 6.40017231 6.39917231
C   CAT1    CAT2    CAT3    CAT4   CAT5    CAT6    CAT7    CAT8   CAT9   CAT10
C    0.0     1.0     1.0     0.0    1.0     0.0     0.0     1.0    0.0     0.0
BLANK card ending node voltage outputs
C  DUM1     DUM2       DUM3       DUM4       DUM5       DUM6       DUM7     DUM8  DUM9     DUM10
C   1.0 3.87400715 -1.5625528 22.4590591 72.4590591 3.31831016 179.999848  3.14159   -4.  -3.14159
C  EAT1       EAT2       EAT3       EAT4       EAT5       EAT6       EAT7       EAT8       EAT9       EAT10
C   4.0 -.99999933    2.30103 .785398163  7.3890561 -.52359878 1.04719755 .813467631 .761594156 .761594156
C    FAT1       FAT2       FAT3       FAT4   FAT5 FAT6       FAT7      FAT8     FAT9   FAT10
C .57734968 1.15470056 .117099841 -.70710678  -3.  5.0 -3.2763323 1.3114543 -4.85987 -.42331
C     GAS1       GAS2       GAS3      GAS4      GAS5    GAS6     GAS7     GAS8     GAS9    GAS10
C  1.15572715 17.2586339 2.48054157 .9990005 .9990005 .9990005 .9990005 .9990005 .9990005  3.999
BLANK card ending plot cards
BEGIN NEW DATA CASE
C     5th of 6 subcases documents pocket calculator applied to many intermediate
C     variables.  This was the beginning of  Orlando Hevia's  R7U.DAT  which was
C     added  6 August 2010.  Exceptional Lists 4 & 26 are close to their minima.
C     The electrical network was substantial,  but is of no interest here so has
C     been omitted.  The  $PARAMETER  data seemed plenty creative to WSM. Volume  
C     is unlimited as  KNT  is unlimited.  But who deserves credit for imagining
C     such use?  Perhaps Prof. Hans Kr. Hoidalen, the author of ATPDraw!  E-mail
C     from Mr. Hevia  dated  Sunday, August 15, 2010  explained:  "The lines are
C     generated by an ATPDraw  option,  @FILE.  It is a sort of pointlist,  with
C     the possibility of one or more columns.  Just before ATPDraw runs ATP,  it
C     generates the lines as you see in r7u.dat. I know the @FILE option because
C     the user sent the question with his problem."   In any case, with such use
C     as this,  the pocket calculator suddenly must handle more voluminous data,
C     particularly intermediate variables.  List 4 (sources) storage of the load
C     flow is being used for the first time as previously-fixed dimensions of
C     the pocket calculator are made variable.  Not all,  but an important part.
C     8 previously-fixed vectors have been made variable.  WSM,  19 August 2010.
NEW LIST SIZES      { Request for a change to program table sizes (VARDIM input)
DEFAULT      3.0    { Any blank field will be converted to 3 times default value
       0       0       0     840      { List 4 = 840 of LISTSIZE.BPA is adequate
BLANK card for lists 11-20.
       0       0       0       0       0     950 { KONST use below requires this   
     390     300     { Optional card for Lists 31, etc. --- as LISTSIZE.BPA
  240000     742     { Offsets for supporting programs (non-simulation overlays)
C   Controls of following request card:   MAXKNT  IOPCVP  NOSTAT { Loop 50 times
POCKET CALCULATOR VARIES PARAMETERS           50       0
C $DEBUG, 9, { For diagnostic output of pocket calculator, activate this request 
C    Following use of tags continues to work,  but is being augmented by the
C    more powerful definition of all such indices at once (see below).
C $PARAMETER   INDOFF=121  KONST=242
C  As long as List 4 is limited to 840 (in LISTSIZE.BPA as of 7 August 2010), it
C  is necessary to optimize the compilation by the  pocket calculator.  This can
C  be seen from the end of diagnostic output,  which shows a need for 913 cells:
C       ----  Assembler after replace any SN= ----
C    117  IP________=IP
C           COPY    IP  IP____      913
C    New input:EOF
C  Without this many cells for List 4, the following line is necessary.  It will
C  reduce the preceding 913 to 789 which is well within the 840 cells available:
C $PARAMETER   NOOPT=0    { Restore compiler optimization that 2nd subcase ended
C $PARAMETER   MAXASS=990  NOOPT=0
C $PARAMETER   MAXLHS=170  ---  For a week or so this was honored, but no longer
C  Having disabled all of the old definitions,  consider the new, better way:
C        1         2         3         4         5         6
C 34567890123456789012345678901234567890123456789012345678901234
C                         INDOFF   KONST   NOOPT  MAXASS  MAXLHS
$PARAMETER LIMITS :          121     242       0     990     170
$PARAMETER
F01001=(KNT.EQ.1.)*229.6111+(KNT.EQ.2.)*233.69775 $$
F01002=(KNT.EQ.3.)*158.12631+(KNT.EQ.4.)*108.91404 $$
F01003=(KNT.EQ.5.)*278.28978+(KNT.EQ.6.)*96.295541 $$
F01004=(KNT.EQ.7.)*239.16153+(KNT.EQ.8.)*187.79161 $$
F01005=(KNT.EQ.9.)*93.658151+(KNT.EQ.10.)*192.55105 $$
F01006=(KNT.EQ.11.)*165.51983+(KNT.EQ.12.)*171.78016 $$
F01007=(KNT.EQ.13.)*269.06742+(KNT.EQ.14.)*256.13115 $$
F01008=(KNT.EQ.15.)*276.15484+(KNT.EQ.16.)*214.35659 $$
F01009=(KNT.EQ.17.)*171.24842+(KNT.EQ.18.)*332.23904 $$
F01010=(KNT.EQ.19.)*341.62624+(KNT.EQ.20.)*72.282156 $$
F01011=(KNT.EQ.21.)*206.20294+(KNT.EQ.22.)*292.75362 $$
F01012=(KNT.EQ.23.)*190.11552+(KNT.EQ.24.)*271.59629 $$
F01013=(KNT.EQ.25.)*278.79531+(KNT.EQ.26.)*312.74695 $$
F01014=(KNT.EQ.27.)*338.04959+(KNT.EQ.28.)*79.377109 $$
F01015=(KNT.EQ.29.)*291.07469+(KNT.EQ.30.)*52.334939 $$
F01016=(KNT.EQ.31.)*113.32174+(KNT.EQ.32.)*198.40889 $$
F01017=(KNT.EQ.33.)*338.27422+(KNT.EQ.34.)*254.87681 $$
F01018=(KNT.EQ.35.)*74.500633+(KNT.EQ.36.)*52.138059 $$
F01019=(KNT.EQ.37.)*44.356605+(KNT.EQ.38.)*301.45684 $$
F01020=(KNT.EQ.39.)*305.09034+(KNT.EQ.40.)*173.39597 $$
F01021=(KNT.EQ.41.)*66.13517+(KNT.EQ.42.)*213.79447 $$
F01022=(KNT.EQ.43.)*83.636753+(KNT.EQ.44.)*242.91365 $$
F01023=(KNT.EQ.45.)*249.7637+(KNT.EQ.46.)*178.4586 $$
F01024=(KNT.EQ.47.)*193.61877+(KNT.EQ.48.)*110.70991 $$
F01025=(KNT.EQ.49.)*221.4351+(KNT.GE.50.)*339.68186 $$
P01001=F01001+F01002+F01003+F01004+F01005+F01006+F01007+F01008+F01009 $$
P01002=P01001+F01010+F01011+F01012+F01013+F01014+F01015+F01016+F01017 $$
P01003=P01002+F01018+F01019+F01020+F01021+F01022+F01023+F01024+F01025 $$
ANGAI=P01003 $$
F02001=+(KNT.EQ.1.)*349.6111+(KNT.EQ.2.)*353.69775 $$
F02002=(KNT.EQ.3.)*278.12631+(KNT.EQ.4.)*228.91404 $$
F02003=(KNT.EQ.5.)*398.28978+(KNT.EQ.6.)*216.29554 $$
F02004=(KNT.EQ.7.)*359.16153+(KNT.EQ.8.)*307.79161 $$
F02005=(KNT.EQ.9.)*213.65815+(KNT.EQ.10.)*312.55105 $$
F02006=(KNT.EQ.11.)*285.51983+(KNT.EQ.12.)*291.78016 $$
F02007=(KNT.EQ.13.)*389.06742+(KNT.EQ.14.)*376.13115 $$
F02008=(KNT.EQ.15.)*396.15484+(KNT.EQ.16.)*334.35659 $$
F02009=(KNT.EQ.17.)*291.24842+(KNT.EQ.18.)*452.23904 $$
F02010=(KNT.EQ.19.)*461.62624+(KNT.EQ.20.)*192.28216 $$
F02011=(KNT.EQ.21.)*326.20294+(KNT.EQ.22.)*412.75362 $$
F02012=(KNT.EQ.23.)*310.11552+(KNT.EQ.24.)*391.59629 $$
F02013=(KNT.EQ.25.)*398.79531+(KNT.EQ.26.)*432.74695 $$
F02014=(KNT.EQ.27.)*458.04959+(KNT.EQ.28.)*199.37711 $$
F02015=(KNT.EQ.29.)*411.07469+(KNT.EQ.30.)*172.33494 $$
F02016=(KNT.EQ.31.)*233.32174+(KNT.EQ.32.)*318.40889 $$
F02017=(KNT.EQ.33.)*458.27422+(KNT.EQ.34.)*374.87681 $$
F02018=(KNT.EQ.35.)*194.50063+(KNT.EQ.36.)*172.13806 $$
F02019=(KNT.EQ.37.)*164.3566+(KNT.EQ.38.)*421.45684 $$
F02020=(KNT.EQ.39.)*425.09034+(KNT.EQ.40.)*293.39597 $$
F02021=(KNT.EQ.41.)*186.13517+(KNT.EQ.42.)*333.79447 $$
F02022=(KNT.EQ.43.)*203.63675+(KNT.EQ.44.)*362.91365 $$
F02023=(KNT.EQ.45.)*369.7637+(KNT.EQ.46.)*298.4586 $$
F02024=(KNT.EQ.47.)*313.61877+(KNT.EQ.48.)*230.70991 $$
F02025=(KNT.EQ.49.)*341.4351+(KNT.GE.50.)*459.68186 $$
P02001=F02001+F02002+F02003+F02004+F02005+F02006+F02007+F02008+F02009 $$
P02002=P02001+F02010+F02011+F02012+F02013+F02014+F02015+F02016+F02017 $$
P02003=P02002+F02018+F02019+F02020+F02021+F02022+F02023+F02024+F02025 $$
ANGBI=P02003 $$
F03001=+(KNT.EQ.1.)*109.6111+(KNT.EQ.2.)*113.69775 $$
F03002=(KNT.EQ.3.)*38.126311+(KNT.EQ.4.)*-11.085959 $$
F03003=(KNT.EQ.5.)*158.28978+(KNT.EQ.6.)*-23.704459 $$
F03004=(KNT.EQ.7.)*119.16153+(KNT.EQ.8.)*67.791614 $$
F03005=(KNT.EQ.9.)*-26.341849+(KNT.EQ.10.)*72.551051 $$
F03006=(KNT.EQ.11.)*45.51983+(KNT.EQ.12.)*51.780155 $$
F03007=(KNT.EQ.13.)*149.06742+(KNT.EQ.14.)*136.13115 $$
F03008=(KNT.EQ.15.)*156.15484+(KNT.EQ.16.)*94.356592 $$
F03009=(KNT.EQ.17.)*51.248423+(KNT.EQ.18.)*212.23904 $$
F03010=(KNT.EQ.19.)*221.62624+(KNT.EQ.20.)*-47.717844 $$
F03011=(KNT.EQ.21.)*86.202936+(KNT.EQ.22.)*172.75362 $$
F03012=(KNT.EQ.23.)*70.115523+(KNT.EQ.24.)*151.59629 $$
F03013=(KNT.EQ.25.)*158.79531+(KNT.EQ.26.)*192.74695 $$
F03014=(KNT.EQ.27.)*218.04959+(KNT.EQ.28.)*-40.622891 $$
F03015=(KNT.EQ.29.)*171.07469+(KNT.EQ.30.)*-67.665061 $$
F03016=(KNT.EQ.31.)*-6.6782645+(KNT.EQ.32.)*78.408894 $$
F03017=(KNT.EQ.33.)*218.27422+(KNT.EQ.34.)*134.87681 $$
F03018=(KNT.EQ.35.)*-45.499367+(KNT.EQ.36.)*-67.861941 $$
F03019=(KNT.EQ.37.)*-75.643395+(KNT.EQ.38.)*181.45684 $$
F03020=(KNT.EQ.39.)*185.09034+(KNT.EQ.40.)*53.395974 $$
F03021=(KNT.EQ.41.)*-53.86483+(KNT.EQ.42.)*93.794469 $$
F03022=(KNT.EQ.43.)*-36.363247+(KNT.EQ.44.)*122.91365 $$
F03023=(KNT.EQ.45.)*129.7637+(KNT.EQ.46.)*58.458596 $$
F03024=(KNT.EQ.47.)*73.618773+(KNT.EQ.48.)*-9.2900913 $$
F03025=(KNT.EQ.49.)*101.4351+(KNT.GE.50.)*219.68186 $$
P03001=F03001+F03002+F03003+F03004+F03005+F03006+F03007+F03008+F03009 $$
P03002=P03001+F03010+F03011+F03012+F03013+F03014+F03015+F03016+F03017 $$
P03003=P03002+F03018+F03019+F03020+F03021+F03022+F03023+F03024+F03025 $$
ANGCI=P03003 $$
F04001=+(KNT.EQ.1.)*65300.+(KNT.EQ.2.)*30667.412 $$
F04002=(KNT.EQ.3.)*31979.935+(KNT.EQ.4.)*81276.349 $$
F04003=(KNT.EQ.5.)*18906.501+(KNT.EQ.6.)*21719.44 $$
F04004=(KNT.EQ.7.)*7124.8541+(KNT.EQ.8.)*4846.9281 $$
F04005=(KNT.EQ.9.)*12418.612+(KNT.EQ.10.)*32587.942 $$
F04006=(KNT.EQ.11.)*14560.237+(KNT.EQ.12.)*6698.0694 $$
F04007=(KNT.EQ.13.)*70803.752+(KNT.EQ.14.)*12315.218 $$
F04008=(KNT.EQ.15.)*24474.203+(KNT.EQ.16.)*24726.58 $$
F04009=(KNT.EQ.17.)*19379.812+(KNT.EQ.18.)*36500.402 $$
F04010=(KNT.EQ.19.)*19104.929+(KNT.EQ.20.)*4845.0075 $$
F04011=(KNT.EQ.21.)*9027.5757+(KNT.EQ.22.)*31403.277 $$
F04012=(KNT.EQ.23.)*7222.6444+(KNT.EQ.24.)*17609.525 $$
F04013=(KNT.EQ.25.)*9836.8418+(KNT.EQ.26.)*7827.2433 $$
F04014=(KNT.EQ.27.)*9097.1448+(KNT.EQ.28.)*16693.945 $$
F04015=(KNT.EQ.29.)*29337.654+(KNT.EQ.30.)*22084.662 $$
F04016=(KNT.EQ.31.)*70211.788+(KNT.EQ.32.)*19456.883 $$
F04017=(KNT.EQ.33.)*5592.8618+(KNT.EQ.34.)*24983.583 $$
F04018=(KNT.EQ.35.)*26539.252+(KNT.EQ.36.)*33746.765 $$
F04019=(KNT.EQ.37.)*40387.027+(KNT.EQ.38.)*86591.522 $$
F04020=(KNT.EQ.39.)*14466.269+(KNT.EQ.40.)*31953.179 $$
F04021=(KNT.EQ.41.)*24428.887+(KNT.EQ.42.)*14013.494 $$
F04022=(KNT.EQ.43.)*200790.43+(KNT.EQ.44.)*19225.061 $$
F04023=(KNT.EQ.45.)*6217.3361+(KNT.EQ.46.)*140243.24 $$
F04024=(KNT.EQ.47.)*34046.067+(KNT.EQ.48.)*8264.9251 $$
F04025=(KNT.EQ.49.)*45450.444+(KNT.GE.50.)*85889.85 $$
P04001=F04001+F04002+F04003+F04004+F04005+F04006+F04007+F04008+F04009 $$
P04002=P04001+F04010+F04011+F04012+F04013+F04014+F04015+F04016+F04017 $$
P04003=P04002+F04018+F04019+F04020+F04021+F04022+F04023+F04024+F04025 $$
IP=P04003 $$
IP________=IP
IQ______=IP
ANGC______=ANGCI
ANGB______=ANGBI
ANGA______=ANGAI
C  Intermediate variable before parameter itself.  |IP=P04003 $$
C  Parameter 1  defined.  Value =  6.530000E+04    |IP________=IP
C  Parameter 2  defined.  Value =  6.530000E+04    |IQ______=IP
C  Parameter 3  defined.  Value =  1.096111E+02    |ANGC______=ANGCI
C  Parameter 4  defined.  Value =  3.496111E+02    |ANGB______=ANGBI
C  Parameter 5  defined.  Value =  2.296111E+02    |ANGA______=ANGAI
BLANK card ends input to $PARAMETER
   5.E-9   6.E-6
     500       1       1       1       1       0       0       1       0
C  The rest of R7U.DAT is of no particular interest, so terminate execution:
$ABORT
BEGIN NEW DATA CASE
C     6th of 6 subcases illustrates a much more efficient representation of
C     the preceding 5th subcase using the new  SELECT  function of the pocket
C     calculator.  The answer is unchanged.  Variable names have been lengthened
C     to avoid warnings,  however.   For the 3 angles,  "ANG"  has become "ANGL"
C     in order that names reach the minimum of 6 bytes.  Finally (the 4th of 4),
C     "IP"  has become  "IPNAME".  Warnings were avoided in the 5th subcase only
C     because the names were internal to the pocket calculator (marked by a "$$"
C     on the right edge).  Using  SELECT,  there is no need for any such special
C     treatment (note that all "$$" have been removed).  WSM.  14 September 2010
C   Controls of following request card:   MAXKNT  IOPCVP  NOSTAT { Loop 50 times
POCKET CALCULATOR VARIES PARAMETERS           50       0
C 34567890123456789012345678901234567890123456789012345678901234
C                         INDOFF   KONST   NOOPT  MAXASS  MAXLHS
C $PARAMETER LIMITS :        121     242       0     990     170
C     Note that the preceding expansion of calculator tables has been removed.
C     Thanks to  SELECT,  storage is minimal.  The numbers themselves do not
C     count as they are stored in available numerical storage.  Rather than 
C     990 lines of assembly language (to which limit  MAXASS  applies),  this
C     data never exceeds 2.  The  SELECT  function itself requires a single
C     line and storage of the value that was selected requires a second.  To
C     prove that default dimensions work,  we reset them:
C 34567890123456789012345678901234567890123456789012345678901234
C                         INDOFF   KONST   NOOPT  MAXASS  MAXLHS
$PARAMETER LIMITS :           60     120       0     270     170
C    First, let's show how simple the use of  SELECT  makes the preceding 5th
C    data subcase.  The same values for  ANGLAI,  ANGLBI,  ANGLCI,  and  IPNAME
C    are produced with great economy of both data and computation as follows:
$PARAMETER
ANGLAI = SELECT ( KNT )
  229.6111 233.69775 158.12631 108.91404 278.28978 96.295541 239.16153 187.79161
 93.658151 192.55105 165.51983 171.78016 269.06742 256.13115 276.15484 214.35659
 171.24842 332.23904 341.62624 72.282156 206.20294 292.75362 190.11552 271.59629
 278.79531 312.74695 338.04959 79.377109 291.07469 52.334939 113.32174 198.40889
 338.27422 254.87681 74.500633 52.138059 44.356605 301.45684 305.09034 173.39597
  66.13517 213.79447 83.636753 242.91365  249.7637  178.4586 193.61877 110.70991
  221.4351 339.68186
ANGLBI = SELECT ( KNT )
  349.6111 353.69775 278.12631 228.91404 398.28978 216.29554 359.16153 307.79161
 213.65815 312.55105 285.51983 291.78016 389.06742 376.13115 396.15484 334.35659
 291.24842 452.23904 461.62624 192.28216 326.20294 412.75362 310.11552 391.59629
 398.79531 432.74695 458.04959 199.37711 411.07469 172.33494 233.32174 318.40889
 458.27422 374.87681 194.50063 172.13806 164.3566  421.45684 425.09034 293.39597
 186.13517 333.79447 203.63675 362.91365 369.7637  298.4586  313.61877 230.70991
 341.4351  459.68186
ANGLCI = SELECT ( KNT )
 109.6111  113.69775 38.126311-11.085959 158.28978-23.704459 119.16153 67.791614
-26.341849 72.551051  45.51983 51.780155 149.06742 136.13115 156.15484 94.356592
 51.248423 212.23904 221.62624-47.717844 86.202936 172.75362 70.115523 151.59629
 158.79531 192.74695 218.04959-40.622891 171.07469-67.665061-6.6782645 78.408894
 218.27422 134.87681-45.499367-67.861941-75.643395 181.45684 185.09034 53.395974
 -53.86483 93.794469-36.363247 122.91365  129.7637 58.458596 73.618773-9.2900913
  101.4351 219.68186
IPNAME = SELECT ( KNT )
    65300. 30667.412 31979.935 81276.349 18906.501  21719.44 7124.8541 4846.9281
 12418.612 32587.942 14560.237 6698.0694 70803.752 12315.218 24474.203  24726.58
 19379.812 36500.402 19104.929 4845.0075 9027.5757 31403.277 7222.6444 17609.525
 9836.8418 7827.2433 9097.1448 16693.945 29337.654 22084.662 70211.788 19456.883
 5592.8618 24983.583 26539.252 33746.765 40387.027 86591.522 14466.269 31953.179
 24428.887 14013.494 200790.43 19225.061 6217.3361 140243.24 34046.067 8264.9251
 45450.444 85889.85
C   Parameter 1  defined.  Value =  2.296111E+02    |ANGLAI = SELECT ( KNT )
C   Parameter 2  defined.  Value =  3.496111E+02    |ANGLBI = SELECT ( KNT )
C   Parameter 3  defined.  Value =  1.096111E+02    |ANGLCI = SELECT ( KNT )
C   Parameter 4  defined.  Value =  6.530000E+04    |IPNAME = SELECT ( KNT )
BLANK card ends input to $PARAMETER
C    That takes care of the 5th data subcase.  Before leaving this 6th subcase,
C    however,  we will illustrate several variations.  First,  the argument of
C    the  SELECT  function,  which is not necessarily  KNT.  Yes, the PCVP loop
C    counter  KNT  is the natural one,  but not the only one.  The argument of
C    SELECT  can involve pocket calculator math as shown by this modification
C    which uses subscript  VALUE4 = LL3 + 1 = 3 + 1 = 4.  The 4th number of the
C    following set is  -11.085959  so this is the value that will be loaded into
C    the left hand side  ANGLCI.  Final detail:  LL3  is an ATP symbol which has
C    integer value 3.  Similarly  LL4  (used later) has value 4.
$PARAMETER
VALUE4 = LL3 + 1.0  $$ { VALUE4 is an intermediate variable so a double dollar
PARAM5 = SELECT ( VALUE4 )    { This param #5 should have value -11.085959
 109.6111  113.69775 38.126311-11.085959 { Truncate points at last one used (4)
C   Also illustrate that comment lines are tolerated within data of SELECT.  The
C   following  PARAM6  for parameter #6  should give a result that is identical
C   to  ANGLAI  when  KNT = 1  (all that is being used):
PARAM6 = SELECT ( KNT )
  229.6111 233.69775 158.12631 108.91404 278.28978 96.295541 239.16153 187.79161
C 93.658151 192.55105 165.51983 171.78016 269.06742 256.13115 276.15484 214.35659
C 171.24842 332.23904 341.62624 72.282156 206.20294 292.75362 190.11552 271.59629
C 278.79531 312.74695 338.04959 79.377109 291.07469 52.334939 113.32174 198.40889
C 338.27422 254.87681 74.500633 52.138059 44.356605 301.45684 305.09034 173.39597
C  66.13517 213.79447 83.636753 242.91365  249.7637  178.4586 193.61877 110.70991
  221.4351 339.68186        {  An in-line comment is legal on partial final card
C   Also illustrate that non-integer subscripts will result in a warning, but
C   not a fatal error.  Integer truncation will be used as in Fortran.  Thus
C   the value of  PARAM7  should be identical to that of  PARAM5.  Once again
C   only 4 data points are required,  but the full first card of 10 will be
C   used to illustrate the following blank card for termination.
VALUE5 = LL4 + 0.5  $$ 
PARAM7 = SELECT ( VALUE5 )    { This param #7 should have value -11.085959
 109.6111  113.69775 38.126311-11.085959 158.28978-23.704459 119.16153 67.791614
BLANK card is needed to terminate SELECT since preceding card was full
C   Finally illustrate that an excessive subscript will result in a warning, but
C   not a fatal error.  Just as for Prof. Hoidalen's data of the 5th subcase,
C   the last data point will be used if a subscript is excessive.  In this
C   case,  the subscript has value 5 whereas only 4 data points are defined.
PARAM8 = SELECT ( LL5 )    { This param #8 should have value -11.085959
 109.6111  113.69775 38.126311-11.085959 
C  Intermediate variable before parameter itself.  |VALUE4 = LL3 + 1.0  $$ { VALUE4 is an intermediate variable so a double dollar
C  Parameter 5  defined.  Value = -1.108596E+01    |PARAM5 = SELECT ( VALUE4 )    { This param #5 should have value -11.085959
C  Parameter 6  defined.  Value =  2.296111E+02    |PARAM6 = SELECT ( KNT )
C  Intermediate variable before parameter itself.  |VALUE5 = LL4 + 0.5  $$
C   ###  Warning.  The argument of  SELECT  is not an integer.  POCKE4  truncates to integer from  4.5000000E+00
C  Parameter 7  defined.  Value = -1.108596E+01    |PARAM7 = SELECT ( VALUE5 )    { This param #7 should have value -11.085959
C   ###  Warning.  SELECT  argument   5  is too large for data.  POCKE4  uses last entry, number   4
C  Parameter 8  defined.  Value = -1.108596E+01    |PARAM8 = SELECT ( LL5 )    { This param #8 should have value -11.085959
BLANK card ends input to $PARAMETER
   5.E-9   6.E-6
     500       1       1       1
C  The rest of R7U.DAT is of no particular interest, so terminate execution:
$ABORT
BEGIN NEW DATA CASE
BLANK