Daniel R Hicks
Reputation
257
Top tag
Next privilege 500 Rep.
Access review queues
 Aug9 awarded Yearling Aug7 awarded Popular Question May29 comment How to add or multiply decimal numbers? The common way of adding decimal numbers is to use "excess-6" arithmetic, where 6 is added if the digit exceeds 9. There are various clever ways to do this, some more "clever" than they should be, but since the operation tends to be highly optimized the cleverness is usually merited. May29 asked What is a common way to measure the “goodness of fit” of an individual data point to a correlation? May23 awarded Teacher Apr3 comment Sheffer stroke the most important advance in logic? It's worth pointing out that, in the parallel world of digital logic design, the NAND gate is considered to be the "universal" logical element -- any digital device you want to build can be built fron NAND gates. (And, in fact, this was often done.) I can imagine that if NAND gates had been invented fairly late in the development of digital electronics (rather than being just about the first thing invented) they would have been considered quite magical. Apr2 comment Pseudo Proofs that are intuitively reasonable Other than the "gives the right answer" part, haven't you defined politics? Mar12 comment Exponential average of logs @RahulNarain -- I've finally gotten the numbers to where they mostly stay positive, so mainly we're handling noise situations and "transitions". But it would be nice to have a general algorithm so I don't have to worry about discontinuities. Mar12 revised Exponential average of logs Corrected algorithm Mar12 asked Exponential average of logs Feb11 accepted peak limiting/audio compression formula needed Feb10 comment peak limiting/audio compression formula needed Put that in an answer and I'll give you the check. Though I went over to DSP and got a suggestion of new_level = 10 * average * tanh(level/(10 * average)), which works pretty well. Feb9 asked peak limiting/audio compression formula needed Feb9 comment What's the simplest curve that has an inflection point at zero and is asymptotic at 100.0? I'd give you another +1 if I could. Came back to this thread seeking an answer to a question similar to the last one but a bit different and your answer gave me the hints I needed. Thanks! Feb7 accepted Sum of averages vs average of sums Feb6 answered Sum of averages vs average of sums Feb5 comment Sum of averages vs average of sums (When I get back to work on Monday I'm going to change the logic to use double-precision on the off-chance that the problem is due to losing precision. But it seems unlikely given the data.) Feb4 comment Sum of averages vs average of sums This is C code, not a spreadsheet, and the numbers are "float" values, so no blanks. Pretty sure it's not a simple bug (though it may well be some sort of conceptual problem). The "sum" of the values in a row is actually accumulated separately from/concurrently with the individual values (though a check shows they add up to equal the sum), so taking the average of the row or some such is not the problem. Feb3 comment Sum of averages vs average of sums (That is, the vertical time dimension is calculated with a moving average. FWIW, the column values tend to differ by a factor of 50 or so, with one column predominating.) Feb3 comment Sum of averages vs average of sums Actually, it's an exponential moving average with N of 50.