A. JUST LOAD BOTH DISKS INTO THE SAME SUBDIRECTORY AND TYPE "WASPS1" OR WASPS2" OR "WASPS3".... "WASPS9"
Q. WHAT IS THE DIFFERENCE BETWEEN WASPS 1 - 9
A. WASPS1-3 ALL EQUAL EACH OTHER WASPS4-6 ALL = EACH OTHER , AND WAPSPS7-9 ALL = EACH OTHER.
THIS IS SO THAT IF ONE GETS DAMAGED THE OTHER TWO ARE STILL INTACT IN THEIR GROUP OF 3 AND WILL ENABLE YOU TO KNOW WHICH ONE IS DAMAGED AND TO BE REPLACED BY A COPY OF ONE OF THE SURVIVORS.
Q. HOW DO I RUN THE TEST FOR SUCH DAMMAGE?
A. JUST TYPE "TESTWASP" AND VIEW THE DISPLAY ON THE SCREEN.
Q. HOW DO I REPLACE THE DAMAGED PROGRAM?
A. IF WASPS4 = WASPS5 BUT WASPS6 = NOBODY, THEN WASPS6 IS THE DAMAGED FILE, JUST TYPE "COPY WASPS5 WASPS6"
Q. HOW DOES THE 1-3 GROUP DIFFER FROM THE 4-6 GROUP AND THE 7-9 GROUP?
A. WHEN YOU RUN THE SIMULATIONS, PROMPTS AND HELP LESSONS APPEAR IN THE DISPLAYS TO GUIDE YOU EASILY THROUGH THE INTERACTIVE GAME-LIKE SIMULATION SESSION. YOU WILL OBSERVE HOW A WASP POPULATION RESPONDS TO VARIOUS ENVIRONMENTS OR SELECTION REGIMENS. A TRITUPLE OF NUMBERS IS A MAJOR PART OF THE DESCRIPTION OF SUCH A REGIMEN. THESE ARE:
(1.) A NUMBER CALLED "W" WHICH IS THE NUMBER OF OFFSPRING A WORKER HAS IF SHE IS SELFISH.
(2.)A NUMBER CALLED "Q1", THE NUMBER OF OFFSPRING THE ROOMMATE (QUEEN) OF A SELFISH WORKER HAS ANYWAY, AND
(3.) A NUMBER CALLED "Q2", THE EXTRA OFFSPRING OVER AND ABOVE THE Q1 PRODUCED BY THE ROOMATE IF THE WORKER IS ALTRUISTIC.
AN ALTRUISTIC WORKER HAS 0 OFFSPRING.
IF I SAY "(W, Q1, Q2) = (3, 8, 9)" I MEAN "W = 3, Q1 = 8, Q2 = 9".
WHEN RUNNING, THE PROGRAM WOULD BE SLOWER IF IT HAD TO RECOMPUTE INTERMEDIATE ARITHMETICAL RESULTS OVER AGAIN IF IT ALREADY HAS A RECORD OF THESE FROM PREVIOUS GENERATIONS. VERSIONS 1 - 3 ARE SMART ENOUGH TO "KNOW" THAT IF YOU CHANGE REGIMENS AND A (W, Q1, Q2) OF (3, 6, 9) IS REPLACED BY A (W, Q1, Q2) OF (3, 6, 9) THAT NOTHING HAS CHANGED AND THE PROGRAM DOES NOT NEED TO RECOMPUTE ITS LOOK UP TABLES.
VERSIONS 4 - 6 HAVE THIS CAPABILITY AS WELL. IN ADDITION THEY RECOGNIZE A REPLACEMENT OF (3, 6, 9) BY (30, 60, 90) AS NO CHANGE AND THE LOOK UP TABLE IS NOT RECALCULATED, SINCE IF YOU MULTIPLY ALL 3 MEMBERS OF THE TRITUPLE BY THE SAME CONSTANT, THE SIMULATION IS UNAFFECTED.
VERSIONS 7 - 9 HAVE ALL THESE CAPABILITIES, BUT ONE MORE DIFFERENCE REMAINS:
THE TRITUPLE (0, 0, 0) WILL LEAD TO EXTINCTION OF THE POPULATION IN ONE GENERATION. VERSIONS 1 - 3 RECOGNIZE THIS WITH AN APPROPRIATE ERROR MESSAGE AND A RETURN TO THE RESTART PHASE.
VERSIONS 4 - 6 SEE (0, 0, 0) AS EQUIVALENT TO THE PREVIOUS TRITUPLE. WHAT EVER IT MIGHT HAVE BEEN SINCE ANY TRITUPLE MULTIPLIED THROUGH BY 0 WILL YIELD (0, 0, 0). TO PUT IT BLUNTLY THE PROGRAM BELIEVES ANY TRIANGLE IS SIMILAR TO A POINT.
VERSIONS 7 - 9, HOWEVER, WILL RECOGNIZE 2 TRITUPLES AS EQUIVALENT IF ONE IS SIMPLY OBTAINABLE BY MULTIPLYING THE OTHER THROUGH BY A CONSTANT, BUT ONLY IF THAT CONSTANT DOES NOT EQUAL ZERO! IF GIVEN (0, 0, 0) THE PROGRAM CRASHES.
Q. HOW DO I KNOW THE RESULTS OF THESE SIMULATIONS ARE MEANINGFUL AND NOT JUST SCIENCE FICTION?
A. THE BASIC LISTINGS END WITH COMMENT LINES CONTAINING A BIBLIOGRAPHY OF REFEREED JOURNAL ARTICLES.
Q. BUT WHAT IF THE PROGRAM GETS CORRUPTED IN STORAGE AND A CONSTANT IN A FORMULA GETS CHANGED?
A. THE TESTWASP COMMAND WILL TAKE CARE OF THAT.
Q. WHEN THE SPEED OPTION IS TURNED OFF, I GET A DISPLAY OF THE "FITNESSES" OF AA, AS, AND SS WASPS. ARE THESE INCLUSIVE OR PERSONAL FITNESSES?
A. THESE ARE PERSONAL FITNESSES. THE CODE FOR THE CALCULATION OF THE INCLUSIVE FITNESSES HAS BEEN LOST, BUT I AM RECONSTRUCTING IT NOW AND IT WILL BE MADE AVAILABLE FOR THESE DISPLAYS IN A FORTHCOMING UPGRADE.
Q. WHAT IS THE PURPOSE OF EDITOR.BAS?
A. FOR MAKING CHANGES IN ALL 9 VERSIONS OF THE PROGRAM SIMULTANEOUSLY BY EDITING THE PARTS THEY HAVE IN COMMON ONCE AND HAVING THE COMPUTER DISTRIBUTE THE CHANGES TO ALL 9 VERSIONS.
ONE CONSOLIDATES ALL WASPS?.BAS VERSIONS INTO ONE FILE CALLED POLISTES BY USING HELBENDR.
ONE APPLIES THE CHANGES TO POLISTES, POSSIBLY WHILE RUNNING EDITOR.BAS.
ALTERNATIVELY, ONE CAN USE GLOBAL.EXE. THE USE OF THESE TWO PROGRAMS IS TO PREVENT THE INTRODUCTION OR DELETION OF CARRIAGE RETURNS AS IT IS ESSENTIAL NOT TO CHANGE THE NUMBER OF LINES IN POLISTES. TO GET THE 9 VERSIONS OF WASPS BACK AGAIN, RUN POLISTES THROUGH THE PROGRAM CALLED MUDPUPPY.
Q. HOW DO I GET AND RUN HELBENDR, GLOBAL AND MUDPUPPY?
A. UNZIP THE SALAMANDERS PACKAGE (CURRENTLY VERSION SAL00000.ZIP ON
SIMTEL UNDER MSDOS/COMPRESS) INTO THE SAME SUBDIRECTORY AS YOUR WASPS PROGRAMS, INCLUDING THE FILE LENGTH. ***WARNING: DO NOT USE SALAMANDERS FROM NORTON COMMANDER!
A. All sources have been moved to archive?.sal in inclusif.zip, and can be extracted
using the salamanders program SAL00000.ZIP (or latest version) in Simtel (e. g. at http://www.coast.net/SimTel/msdos/compress.html)
CURSOR ADRK5A0S
Q. Are the fitnesses displayed on the screen and labled "FITNESS", personal fitnesses or inclusive fitnesses?
A. The original WASPS programs display personal fitnesses. If you replace the "P" (for "personal") in WASPS with "C", as in "WASCS3.BAS", instead of "WASPS3.BAS" then you get a program in which Corrected Inclusive Fitnesses are used, instead. Replacing the "C" or "P" with a "U" will get you a program in which uncorrected inclusive fitnesses are used.
Q. What are these different kinds of fitness anyway?
A. The Personal Fitness is the number of own begotten offspring an individual contributes to the next generation regardless who rears, rescues, or provides for them. This implies if a non-parental provider, like a prospective aunt of a potential offspring, shows up and offers service, the prospective parents will conceive the offspring, or the offspring will be saved from an early death, because the parents are already producing the maximum they can provide for. An individual's Uncorrected Inclusive Fitness is the number of "offspring equivalents" the individual provided for, thus enabling their existence, regardless who their parents are. The variable, RHO, the number of offspring worth one niece or nephew when measured between siblings as in this model, is in this case the number of offspring equivalents in a niece or nephew (a fraction of course). Every time the worker rears a niece or nephew she rears RHO offspring equivalents and (1 - RHO) stranger equivalents. The Corrected Inclusive Fitness equals the Uncorrected Inclusive Fitness plus C, where C is the individual's "Group Karma". C = the total number of stranger equivalents reared this generation divided by the population size. Since the population size has been normalized to 1 (the vector, which gives the numbers of the different kinds of colonies, has been converted to give the frequencies of the kinds of colonies instead) the numerator of C is excluded in the computer programs' source code as it is implicitly equal to one.
Q. How do we know these numbers weren't pulled out of a hat?
A. If you replace the "A" in any of the WAS?S?.BAS versions with an "O" as in "WOSPS1.BAS" the program will be one which allows the user to test the validity of the fitnesses. Instead, you can read the algebra and discussion in my paper in the Journal of Theoretical Biology (1979) v81 pp577 - 586.
To test the Personal and Inclusive Fitnesses displayed by these programs you want to substitute the name of the WASPS version you wish to run for the word "TARGET" into the BATch file called "DISPLAY.BAT".
This is done with a call to GLOBAL and then typing "DISPLAY".
Assuming we are in the D drive in a subdirectory called "SUB", and that our GWBASIC-like BASIC is called "GBASIC", we get this screen display while working:
D:\SUB>global
replaced string TARGET
replacing stringWOSPS1
FILE IN? DISPLAY.BAT
D:\SUB>display
GBASIC WOSPS1 < WASPDATA > T
GBASIC < DSPLDATA
1 File(s) copied
D:\SUB>
.
After the call to GLOBAL just type DISPLAY, and the batch file will generate a table of the gene frequency as a function of time for you. This will represent the result of 2 simulations. In the first, the kin selection model runs. In the second, a classical model, with no social interactions runs. The fitnesses, which most times change each generation, of the 3 genotypes are looked up from a table, which was laid down in the previous run, and used directly. The table is filled by recording the fitnesses from the first run which is a normal "kin selection" (i. e. with social interactions) run.
Q. How do I understand these results?
A. The file P contains the gene frequencies. The 1st 13 lines represent the 1st run whereas the 2nd 13 lines represent the 2nd run. If all goes well, the 2nd 13 gene frequencies should equal the 1st 13 ones. This will be true whether Corrected or Personal Fitnesses are used.
Q. What about Uncorrected Inclusive Fitness?
A. Alas, the use of uncorrected inclusive fitness will not enable the classical run to track the kin selection run.
Q. Does this mean that Uncorrected Inclusive Fitnesses are wrong?
A. No. They are historically and scientifically meaningful. This is because William D. Hamilton, who discovered Kin Theory and Inclusive Fitness, was looking for a measure of fitness which could never decrease as a function of time.
After all, R. A. Fisher's Fundamental Theorem of Natural Selection (which says the rate of change in mean fitness = the variance in fitness) as a corollary predicts that fitness can never decrease as a funcion of time. This, so the reasoning goes, is because a variance being a square of another number cannot be negative. However, Corrected Inclusive Fitness and Personal Fitness can decrease with time (see below). But Hamilton in his 1964 model assumed half-dominance, i. e. a worker with one altruism gene and one selfishness gene spends half of her resources working for herself and half for her queen. Under these conditions Uncorrected Inclusive Fitness will never decrease as a function of time.
Q. Where do I find the results if I run DISPLAY.BAT?
A. As stated earlier, the file, P, contains the gene frequencies alone. The rest of the results will be saved in the file called "T".
Q. When does fitness decrease as time goes on?
A. The mean Corrected Inclusive Fitness and mean Personal Fitness equal each other. These means decrease with time when the grandparents with the fewest grandchildren are being selected for, and by extension the parents with the fewest offspring are being selected for. This happens when there is a "parent-offspring conflict" but the offspring are winning.
Q. How can that happen?
A. Easy! If giving up 1 of her own offspring lands a worker 1.001 nieces and/or nephews (i. e. giving up 1000 own offspring lands her 1001 nieces and/or nephews) it is good for her parents but not for her. There are insufficient offspring equivalents in 1001 of your sib's kids to compensate you for the loss of 1000 of your own potential kids. However, if you get your way because your parents are gone with the last generation, you as a worker get selected for if you produce the 1000 of your own kids. Thus your parents loose 1001 potential grandchildren to gain 1000 actual ones.
In short, Personal Fitness and Corrected Inclusive Fitness decrease if and only if 1 < q2/w < 1/RHO, where, an altruistic worker is said to give up W own offspring when her queen is having Q1 anyway but gains Q2 extra because of the help.
Uncorrected Inclusive Fitness will decrease if and only if the more dominant of the pair (altruism / selfishness) is being selected for. By dominance is meant "A worker with one A gene and one S gene spends more or less than (but not exactly) 1/2 of her time, and other resources, altruistically.". It is assumed that the fraction chosen" by an AS worker will be the same for all AS workers. "Altruism dominant" would imply more than 1/2 of the resources are being used altruistically.
Q. How do I get all these WOSCS1.BAS and WASUS4.BAS programs anyway?
A. Unzip SAL00000.ZIP (found on SimTel under COMPRESSion (It's short for SALAMANDERS.) - WASPS users are absolved from the SALAMANDERS shareware fee) into the same subdirectory as WASPS. Then type:
MUDPUPPY and type ARCHIVE1.SAL at the prompt. It will generate the 37 BASIC files. Have .9 MEG available for them. Make sure the file named "LENGTH" has a 50 in it.
If LENGTH has 6, call GLOBAL and give it 6, 50, and "LENGTH" as inputs.
Q. What about EXE files for all those new BAS files?
A. If I'm lucky you'll get uuencoded versions of them. Otherwise, you will have to compile them or run them under interpreted BASIC.
Q. Why did you do it this way instead of one program with lots of IF statements?
A. I want to demonstrate a realistic ratio between the run times on the program with vs. without the optimizer (speed option) and IF statements in the innermost loop would have confounded the meaning of the perceived difference in run times.
BATCH FILES UPDATE
TESTWASP.BAT and COPYWASP.BAT, which have been described in WASPS.DOC have their TESTWASC.BAT, COPYWOSU.BAT, etc. counterparts. The last "S" is replaced by a "B" as in TESTWOSB.BAT if you only wish to test BAS files. An "X" as in TESTWOXU.BAT" is employed if you only wish to test EXE files.
