The story of individual 8 from generation 18. (draft)

Once, there was a curved bicone named individual 8 who was from generation 18. In many ways, it was similar to many other bicones but in fact, this individual was the best bicone that ever lived with a fitness score of 5.22097. It achieved this by having the following parameters; an inner radius of 6.31483, and 3.97024; lengths of  77.6017 and 40.9244; quadratic coefficients of, 0.00260615 and, -0.0103313; and finally, a linear coefficient of -0.197494 and

0.36119. Unfortunately, individual 8’s lineage ended with it as it never met that special someone (crossover) and didn't survive into the next generation (reproduction) despite having a 77.7% of doing at least one of these. 

This bicone had a sibling with a fitness Score of 4.65861 and had the following parameters: 6.05614,79.663,-0.000353038,0.0331969 (side 1)

1.92878,37.9849,-0.00213265,0.156724  (side 2).

Unfortunately, this bicone also suffered the same fate as its sibling and failed to leave a lasting mark on this hypothetical world. 

These two individuals had parents from generation 17 whose name’s were individuals 39 and individual 8. After having individuals 8 and 9 in generation 18, individuals 39 and 8 from generation had a nasty divorce (possibly leading to 8 and 9’s aversion to marriage) and remarried. In these marriages, they both produced two more children that would be our previous antenna’s step-siblings. 

Antenna 8 from gen 17 married antenna 46 to produce antenna 20 with a fitness score of 4.57122,  and antenna 21 with a fitness score of 4.71056. Individual 20 shared the same genes for the second set of linear and quadratic coefficients as its more successful step-sibling individual 8. 21 had no similarities to individual 8 but had 4 children in generation 19. 

Antenna 39 from gen 17 married antenna 5 to produce antenna 28 with a fitness score of 4.71808 and antenna 29 with a fitness score of 4.90119. Antenna 28 shared had an identical side 1 to individual 8 and also shared the same inner radius and length on the second side, and had 2 children in generation 19. Individual 29 had no similarities to individual 8 but really got around and had 8 children with various partners in generation 19. 

This run followed the first symmetric run. It had an asymmetric length, but the opening angle and inner radius were kept symmetric. 

10 individuals were evolved over 17 generations using 100,000 neutrinos. All selection was done through roulette, and all individuals were formed through crossover and the old mutation method (where individual genes had a 40% to be mutated, and mutated genes were chosen from a gaussian based on the mean and standard deviation of that gene's value in the generation).

This is considered to be the first run of "good" data. Prior to this run, we ran with few neutrinos and had errors which needed to be resolved in AraSim and with the way we handled XFdtd's simulation results being passed to AraSim. Those were resolved in the summer before this run.

This was a symmetric run of 10 individuals over 15 generations using 100,000 neutrinos per individual. This preceded the substantial modifications made by Ryan to the GA. All individuals were formed through crossover, but there was a probability of 40% for genes to be mutated. In this case, mutation was a complete change to the gene, though it was based on the average value of that gene in the generation. There was no reproduction or mutation. All individuals were selected through roulette.

Green individuals indicate that there was no penalty applied, while red individuals had a penalty factor multiplied by their effective volumes. The penalty was of the form exp(-(R-7.5)^2), where R is the outer radius, in cm, of the bicone.

This run was started on 10/23/2020. The purpose was to attempt to demonstrate evolution by beginning from 50 identical individuals in the initial generation (which had previously produced a low score).

This run was the first time we removed the penalty on fitness scores for antennas which exceeded the borehole radius. It was also the first time we increased the number of neutrinos thrown up to 300,000. 50 fully asymmetric individuals were evolved over 25 generations.

There was a 50/50 split for roulette/tournament selection and 75%/10%/15% for crossover/reproduction/immigration. While the evolution was somewhat flat, we do believe we demonstrated evolution because the average score rose as the run evolved. 

(Note that the final generation was interrupted but the plot was still made, hence the drop to 0 for all scores on the plot)

Beam patterns with ARA bicone contained in a cylinder of air surrounded by a shell of ice.

 

Slides contatining my notes on matching circuits.

https://docs.google.com/presentation/d/1x25nhiqaW7LvPZ1pNZ5O4ZzsWZbtgqxBQ5haB9uWgQY/edit?usp=sharing

This run was a short run (stopped due to resource limits) that was conducted to test the minimum length constraint that had been applied in the past. Previously, the minimum length was set to be 37.5 cm for each cone; in this run, it was lowered to 10 cm. It ran for 60 generations with 50 individuals per generation, each evaluated with 300,000 neutrinos. To compare to the paper run, this run was performed using the asymmetric algorithm, meaning each individual had 6 genes (length, inner radius, and opening angle for each cone). The initial generation began with identical individuals, with genes matching the genes of the best performing individual mentioned in the GENETIS paper.

The ratio of generative operators is as follows: 72% crossover, 22% immigration, 6% reproduction. Mutation was also used, with 1% of genes produced by crossover mutated by adding a number chosen from a Gaussian centered at 0 with a width of 5% of the gene's value.

The ratio of selection operators was: 20% roulette, 80% tournament, 0% rank, 0% elite.

Attached are the run details, violin plot, and "rainbow" plot.

This is the run that is discussed in the GENETIS paper submitted to Physical Review D in December of 2021. It was conducted beginning on March 19, 2021. The preprint can be found here. It ran for 31 generations with 50 individuals per generation. Each individual was run through AraSim for 300,000 neutrinos.

The ratio of generative operators was: 72% crossover, 22% immigration, 6% reproduction. 

The ratio of selection operators was: 80% roulette, 20% tournament, 0% rank.

The highest scoring individual had a fitness score of 5.24. However, when it was rerun with many more neutrinos (3*10^7), it had a score of 4.90, an 11% improvement over the ARA bicone.

This run was conducted before the introduction of the automatic run_details.txt generator. Instead, there is a file called Run_Notes.txt (attached) which details some of the errors that were encountered. These errors were remedied during runs by removing the offending AraOut file (which had too high of a fitness score due to a weight being calculated to be greater than 1, which shouldn't be possible) and have since been prevented by modifying the data file AraSim uses to calculate the weights.

This run was started on February 2, 2022. It is a full run of 50 generations with 50 individuals per generation using the quadratic version of the loop. This means that each individual is defined by 8 genes (length, inner radius, linear coefficient, and quadratic coefficient for each cone). Attached is the txt file run_details.txt that is automatically generated when the loop is run. Each individual was run for 300,000 neutrinos.

This run used the usual ratio of generative operators: 72% crossover, 22% immigration, and 6% reproduction. It also used the proper mutation function: 1% of genes created through crossover were mutated by adding a number chosen from a Gaussian distribution centered at 0 with a width of 5% of the gene's value (MUTATION WAS NOT USED BUT WAS AVAILABLE).

This run was the first full run in which the rank selection operator was used. The ratio of selection operators was: 0% roulette, 10% tournament, 90% rank, 0% elite. 

This run used the script fitness_check.py to average the scores of identical individuals that appeared across multiple generations. This gives us a more accurate measure of those individuals' scores. It may also explain why this run appears so flat (on the violin plot): fluctuations in the scores should die down as repeated individuals have more accurate scores, and newly generated individuals that perform highly regress to their mean (actual) score when they are reproduced/recreated through crossover. This has led us to question our GA parameters, as Audrey and Autumn demonstrated that roughly 1/4 individuals in the run are repeat individuals.

This is a duplicate post of a previous post from the end of 2020 where I listed the important runs with some of their details in a table (as below). I am extending this table to include the important runs that have been conducted since this post. This includes the run used for the paper as well as the curved run done earlier this year.

In order to access the data for these runs, you can find them by going to this directory: /fs/ess/PAS1960/BiconeEvolutionOSC/BiconeEvolution/current_antenna_evo_build/XF_Loop/Evolutionary_Loop/Run_Outputs

Some of these runs can also be accessed in the old project space directory, though they should all be contained in the above directory. Here's the path if interested: 

The runs are contained in directories available in the above path. Use caution when listing files in some of these directories--some contain many files (primarily the .uan files -- more recent runs are better organized), which means it may take a long time to list files/directories.

Bellow lies plots testing different scores and comparing them using a chi^2 score.

The functions used are as follows

Gaussian: e^(-2) (Red)

Inverse: 1/(1+(O-E)^2) (Purple)

Algebraic: 1/(1+(chi^2) )^(3/2)) (Green)

Chi: 1/(1+chi^2) (Blue)

Which are plotted here:

I'm making this entry so that I can record some interesting papers we find on genetic algorithms. Feel free to update this list with links to papers and maybe make a description of what was interesting/of note in the paper.
 

• Run Type
  • AREA
• Run Date
  • 04/19/22
• Run Name
  • 20220419fahimi5run1
• Why are we doing this run?
  • Check to see if stringReplacement2.py changes were successfull
• What is different about this run from the last?
  • added below code in order to create a fitness file inside of each generation to display the fitness scores for just that generation, as opposed to just having one big file full of all the generations.
    f2 = open(source + "/" + "gen_{}".format(gen) + "/fitnessFile_gen_{}".format(gen) + ".txt", 'a')
    f2.write(','.join(mean_Veff_array) + "\n")
    f2.close()
• Symmetric, asymmetric, linear, nonlinear?
  • N/A (AREA run)
• Number of individuals (NPOP)
  • 12 individuals
• Operators / Selection methods used (% of each)
  • roulette crossover 50%
  • roulette mutation 16%
  • tournament crossover 18%
  • tournament mutation 16%
• Are we using the database?
  • N/A (AREA run)

We are trying to do a short run of the asymmetric bicone so that we can see how it tries to evolve the antenna when the minimum length is lowered from 37.5 cm to 10 cm. Currently, there is a problem with the loop in the asymmetric version. Attached is the run detail file.

• Run Type
  • AREA
• Run Date
  • 04/05/2022
• Run Name
  • 20220405fahimi5run2
• Why are we doing this run?
  • to see if the linear dependence on frequency was properly implemented
• What is different about this run from the last?
  • Under closer examination, the previous run of AREA appeared to not linearly depend on frequency, so we recompiled the GA in order to fix this
• Symmetric, asymmetric, linear, nonlinear?
  • N/A (AREA run)
• Number of individuals (NPOP)
  • 50 individuals
• Number of neutrinos
  • 10,000 per seed
    • 4 seeds per individual
• Operators / Selection methods used (% of each)
  • roulette crossover 50%
  • roulette mutation 16%
  • tournament crossover 18%
  • tournament mutation 16%
• Are we using the database?
  • N/A (AREA run)
• Result
  • There is a problem with how the GA has assigned the Veff values, resulting in many individuals being assigned zero Veff, however the linear dependence seems to have been implemented successfully

Note: We also ran a run called 20220405fahimi5run1, which was a short test run with a very low amount of individuals and neutrinos which seemed to work fine; as in the linear dependence seemed to work.

https://docs.google.com/spreadsheets/d/1vvcmjByKfcns0-tbAjtePB8ZVGsXAKxXCfbMc99weMI/edit?usp=sharing Here is the spreadsheet link for the population test. 

The reviewer for the paper we recently submitted mentioned that our step sizes at which we are measuring the gain in XFdtd may be too large. They pointed out that there appear to be "lobes" at 400 MHz. I ran the antenna we presented in the paper through XFdtd using a step size of 5 degrees (what we've been doing) and a step size of 1 degree (the reveiwer's recommendation). Attached are antenna responses for these two different step sizes at 200 MHz and 400 MHz. Qualitatively, there are noticeable differences in the "jaggedness" of the 400 MHz plot and how extreme the minima and maxima appear, but the basic shape is preserved. We can try to do a more quantitative analysis (ex: run these through AraSim), but doing so may be more time intensive than necessary considering that AraSim may require 5 degree steps and that adding these images to an appendix may be sufficient anyway.

The reviewer of the paper we are trying to publish (as well as other external colleagues we showed the paper to) asked about our minimum length we implemented in the loop. Currently, we cut off the length at a minimum of 37.5 cm for each cone (minimum of 75 cm for the full cone). There is an ELOG post from Amy that presented the reasoning for this ( http://radiorm.physics.ohio-state.edu/elog/GENETIS/81 ). We initially thought that we needed a minimum because the antennas were evolving to be small (around and lower than the minimum). We thought that XFdtd was being inaccurate at lengths lower than 37.5 cm, so we set that as a minimum. To see if we can replicate any strange behavior from XFdtd, I simulated antennas at and below our minimum length and generated plots of their antennas responses. The genes are listed below, corresponding to the patterns in the plot. The results do not look unusual (that is, not dissimilar to the other bicones we have generated, including the ARA bicone) and seem to show an improved sensitivity in the upward direction for shorter bicones. The equation we used to arrive at this minimum (f = c/(4L)) might be indicative of a maximum length rather than a minimum.

2.4892,37.5,-0.00142604,0.032832
4.64522,37.5,0.00153863,-0.14004
2.4892,34.5,-0.00142604,0.032832
4.64522,34.5,0.00153863,-0.14004
2.4892,31.5,-0.00142604,0.032832
4.64522,31.5,0.00153863,-0.14004
2.4892,27.5,-0.00142604,0.032832
4.64522,27.5,0.00153863,-0.14004
2.4892,23.5,-0.00142604,0.032832
4.64522,23.5,0.00153863,-0.14004
2.4892,19.5,-0.00142604,0.032832
4.64522,19.5,0.00153863,-0.14004
2.4892,15.5,-0.00142604,0.032832
4.64522,15.5,0.00153863,-0.14004
 

• Run Type
  • AREA
• Run Date
  • 01/18/22
• Run Name
  • 20220118fahimi5run1
• Why are we doing this run?
  • to see if the linear dependence on frequency was properly implemented
• What is different about this run from the last?
  • attempted to evolve gain pattern as a linear function of frequency
• Symmetric, asymmetric, linear, nonlinear?
  • N/A (AREA run)
• Number of individuals (NPOP)
  • 50 individuals
• Operators / Selection methods used (% of each)
  • roulette crossover 50%
  • roulette mutation 16%
  • tournament crossover 18%
  • tournament mutation 16%
• Are we using the database?
  • N/A (AREA run)