Recent content by MarcJ

I have cleaned up the Python code and created some plots for the values of ##a##, ##b## and ##c## mentioned above. Noise is simply added to the 'ideal' data, has a mean of 0 and a standard deviation of 0, 1 and 10. What I infer is that, at higher noise levels, the mean of the residual is...

I'm getting closer, but we're not there yet. Let me first reply to your thoughts. We don't know the exact distance ##d## in the system, but we can make a well-defined move (##\Delta d##). That's what I meant when I said we don't know our starting point on the curve. So we need ##b## to account...

Thanks Doctor! Thing is, I'm not using Excel of even high-level libraries. This runs on an embedded platform where memory is tight. All code is in C. So even for the linear regression we're using now, I have to spell out the equations and make sure nothing overflows. That works quite well, but...

Hello all, I found this thread because I'm trying to do the same thing: measure a distance using a capacitive sensor. As an extra, I'm doing this as the data comes in, so a streaming algorithm would be best. Normal linear regression can easily be used with streaming data, since only summations...