Divide by Zero Error: Solving Pendulum Radius

[Sam Smith]

I have data from a pendulum and I am using it to work out the radius of the pendulum. I have acceleration in the x and y directions and so thought this would be easy enough. Simply I determine the (velocity in the x direction)^2/acceleration in the y direction. However when I use python to give me a graph I notice that I get a graph with a large peak at the extremes (ie where the pendulum monetarily stops at the extremes and I am therefore dividing by zero) Any way around this?

 

[jbriggs444]

By "x" and "y", one assumes that this amounts to "tangential" and "radial" for an accelerometer mounted on the pendulum. Your complaint is that at the ends of the pendulum's arc you have [near] zero tangential velocity and [near] zero radial acceleration. You suffer from a loss of significance because the error bounds on the two values are as large as the quantities themselves.

One possibility is obvious. Filter your data to use the figures where x velocity is largest.

 

[Sam Smith]

Yes I took this approach taking instantaneous readings. It was correct for some data but not another set. I am hoping to cross check them now so that I can be sure I can always get these values correct