Distribution of simulated data
The following study was made in order to understand the behaviour of Kolmogorov-Smirnov and non-exponential count statistics.
The input data for each simulated bin consisted of 563 values that were computed as sum of squares of two numbers obtained from randn function from Octave (Matlab).
Each run consisted of analysis of 180 simulated bins with KS statistic and determining whether it exceeds the critical value (0.056934) for alpha=0.05 confidence level.
The parameter lambda of the exponential distribution was either determined using median or hardcoded as 0.5.
