Search code examples
c++matlabcurvestatistics

Compare plots analytically / statistically


I created some plots of x1 and x2 with respect to y. My questions are:

  1. How can I compare Plot B with the reference data (Plot A)? What kind of analysis can I do to find out how well they match analytically or statistically? (Is there any algorithm both in Matlab and C++ that I can utilize? I have checked: Compare Plots in matlab)
  2. Instead of comparing the data separately- x1 to x1, x2 to x2 in each plot, is the any way to combine x1 and x2 first, then do the comparison?

I'm using simple sine and cosine plot as an example below but the plots can be any kind of curves. enter image description here

In opencv, we can compare histogram and compute the matching using these metrics: Correlation, Chi-Square, Intersection, Bhattacharyya distance (http://docs.opencv.org/doc/tutorials/imgproc/histograms/histogram_comparison/histogram_comparison.html)


Solution

  • You have to come up with some precise definition of "compare analytically/statistically". The term is too vague. If you wan't to compare anything, you need to define some criterions. I.e. maximal value of function in given interval, average area below the graph and so on.

    I think, you also don't want to analyze plots but rather the plotted data.