diff options
Diffstat (limited to 'benchmarks/dcn19.dat')
| -rw-r--r-- | benchmarks/dcn19.dat | 763 |
1 files changed, 763 insertions, 0 deletions
diff --git a/benchmarks/dcn19.dat b/benchmarks/dcn19.dat new file mode 100644 index 0000000..ec2b2ea --- /dev/null +++ b/benchmarks/dcn19.dat @@ -0,0 +1,763 @@ +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
|