Interesting math fact of the day #198:
where F(n) is the nth Fibonacci number.
seen from Czechia

seen from United States

seen from United States
seen from China
seen from Germany
seen from United States
seen from United States
seen from Belgium

seen from United Kingdom
seen from China

seen from Sweden

seen from Maldives

seen from Malaysia
seen from Kenya

seen from Sweden

seen from Australia
seen from Malaysia

seen from United States
seen from Netherlands
seen from Kosovo
Interesting math fact of the day #198:
where F(n) is the nth Fibonacci number.
Matrix Multiplication - Ex. 2
Patreon
I'm sure there's a reasonable explanation for it at the higher levels of math, but why is basic matrix multiplication Like That.
Patreon | Ko-fi
Patreon | Ko-fi
Patreon | Ko-fi
Patreon | Ko-fi
A big shout-out to Volker Strassen and Shmuel Winograd for showing us more efficient ways to multiply big matrices!
I finally got around to writing a unit test for my implementation of Strassen's algorithm and tested it on a couple 8×8 matrices. Some of the numbers were good, but the lower-left quadrant of the result was way out of whack. Turns out I had the wrong order of matrix multiplication for calculating that quadrant, and once fixed, everything agreed to within an arbitrary 𝜀.
So now it works as it should, but timings reveal that for 8×8 multiplicands, at least, the naïve algorithm is much faster. I'm not sure what Nthreshold is for my implementation, but I'm guessing the real benefits would show themselves starting at maybe 128×128.