1,126 reputation
815
bio website embeddedrelated.com/blogs-1/…
location Arizona
age
visits member for 4 years, 4 months
seen Nov 24 at 20:00

Nov
9
comment Simple numerical methods for calculating the digits of $\pi$
that's easy: since the arctangent is a power series, convergence is dictated by the largest argument to the arctangent. In my case, it's 2/11.
Oct
15
comment How to move a body to a point using only velocity?
NJGUY: you still need to clarify your question. What do you mean by "move" ? Are you writing software to display something at certain points in time? Or do you just want to know the velocity (vx, vy) required to get an object from point (x1,y1) to point (x2,y2) in some time T?
Oct
15
comment How to move a body to a point using only velocity?
The writeup in Wikipedia of Bresenham's algorithm seems more complicated than it should, and also seems like it does not handle lines with steep slopes very well. This really shouldn't be that complicated. We also don't know whether NJGUY wants to draw individual pixels (in which case Bresenham's algorithm is a good match) or just wants to calculate an object's position at various positions in time. Whether you call it vector algebra or two parametric equations x = fx(t), y = fy(t), I still think it's simpler than trying to use time-independent algorithms.
Oct
15
comment How to move a body to a point using only velocity?
@RobertHarvey: no, you don't need the slope if you are using vector algebra.
Oct
15
comment How to move a body to a point using only velocity?
Please clarify your question. Is the second body's position fixed?
Mar
28
comment What comes after tetration ? And after ? And after ? etc.
the symbols didn't get displayed properly
Mar
20
comment Reed-Solomon encoding in GF2?
and there's no relation between GF(256) and the Galois fields used in LFSRs? (I can never remember the technical term for those, a ring over a characteristic polynomial of degree N mod GF(2), or something like that)
Mar
20
comment Reed-Solomon encoding in GF2?
oh, I see, so you can't arbitrarily choose the pair (255,239), for GF(256) it has to be (256 +/- 1, m).
Feb
23
comment Is it possible to build a circle with quadratic Bézier curves?
just curious: what software did you use to draw the diagram?
Feb
10
comment Calculating characteristic polynomials of matrices in GF(2)
So you have to calculate the determinant?
Jan
8
comment simple example of recursive least squares (RLS)
That's helpful for some of the conceptual understanding, but how do I figure out the K's? And I still really need to see some numerical data to understand how to choose the gain parameters in practice. Does RLS also give you a global error/noise estimate?
Oct
7
comment comparing bit lengths of binary numbers
thanks! It looks correct, I'll accept as soon as I have a chance to read more carefully. (Proofs have always eluded me. :/ )
Oct
7
comment comparing bit lengths of binary numbers
I didn't say I would get random results, but I couldn't find a counterexample.
Oct
3
comment Understanding Primitive Polynomials in GF(2)?
thanks for the reference! I like Saxena & McCluskey's algorithm for finding primitive polynomials.
Oct
3
comment Extended Euclidean Algorithm in bit representation problem
Thanks for mentioning Blankinship's algorithm! I'd never heard of it, but was able to implement it in Python for GF2 very easily.
Aug
10
comment generating a random periodic function with bounded amplitude and bounded fourier coefficients
re: randomly generating + scaling -- that's about what I thought of, but it skews the resulting probability distribution of the coefficients.
Aug
10
comment generating a random periodic function with bounded amplitude and bounded fourier coefficients
yes, a computational sense, let me elaborate; that's too loose of a bound.
Feb
11
comment convolution square root of uniform distribution
hmm, upon further reflection it seems like there is no such pdf; the convolution of f(x) with itself would always have a maximum when it lines up.
Feb
11
comment numerical integration for N datapoints
Huh, never heard of Romberg integration before, thanks... What if h is fixed (i.e. velocity measurements made once every N milliseconds) rather than something that can be increased adaptively?
Feb
11
comment numerical integration for N datapoints
Well, I understand that part when N is small (I guess it's kind of like windowing functions in FFTs), but not when N is large.