-
Computer generated random numbers are actually pseudo random numbers, actually generated by an algorithm
-
This algorithm does generate varying numbers, but some are repeated more than others
-
Works better for larger number ranges, for small ranges not all numbers are generated
-
Important: This algorithm IS predictable and not meant for actual use. It will generate varying numbers that may appear to be random, but the are predictable.
-
Also, I have written prototype algorithm that will randomize the list using the same algorithm rather than shuffling it, which is less predictable, see stats below
- Makes a list of all the integers in the given range
- Shuffles the list a few times by pushing all the numbers down by 1. The last number is moved to the start. Every 5 number generations, the program will push the list a few times.
- Picks an algorithm determined number from the list
.gif)
.gif)
- As shown, the both our algorithm's and our prototype algorithm's coverage times are quite comparable to the random module's, but still are far from the theoretical best
- Like those generated by the random module, the numbers generated are scattered
- Though the values are mostly scattered, some things should be noted
- All values will eventually appear, since the list is being shuffled, but some will appear more than others, and it will take time for all of the numbers to appear
- Distinct parabolic shapes can be observed left-top, middle-top, and middle-right.
- Unlike our algorithm's generations, no significant and reeating pattern can be observed in the random module's generations
- Testing is undergoing to figure out the most efficient number of shuffles per cycle. See a clip below.
.gif)
- Our Prototype algorithm randomizes the list using the same random picking algorithm rather than shuffling it.
- Is less predictable than original algorithm and has few observable patterns.
- It should be noted that prototype algorithm is much slower because when it randomizes it has to do a full scan, search, and list append.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent