Module trigger_fits

Maximum likelihood fit for the coefficient alpha for a distribution of discrete values p(x) = alpha exp(-alpha (x-x_t)) above a threshold x_t.

vals: sequence of real numbers none of which lies below thresh thresh: threshold used in the fitting formula

Maximum likelihood fit for the coefficient alpha for a distribution of discrete values p(x) = alpha x exp(-alpha (x**2-x_t**2)/2) above a threshold x_t.

vals: sequence of real numbers none of which lies below thresh thresh: threshold used in the fitting formula

Maximum likelihood fit for the coefficient alpha for a distribution of discrete values p(x) = ((alpha-1)/x_t) (x/x_t)**-alpha above a threshold x_t.

vals: sequence of real numbers none of which lies below thresh thresh: threshold used in the fitting formula

The fitted exponential function normalized to 1 above threshold

xvals: the values at which the fit PDF is to be evaluated alpha: the fitted exponent factor thresh: threshold value for the fitting process

To normalize to a given total count multiply by the count.

The integral of the exponential fit above a given value (reverse CDF) normalized to 1 above threshold

To normalize to a given total count multiply by the count.

The fitted Rayleigh function normalized to 1 above threshold

xvals: the values at which the fit PDF is to be evaluated alpha: the fitted exponent factor thresh: threshold value for the fitting process

To normalize to a given total count multiply by the count.

The integral of the Rayleigh fit above the x-values given (reverse CDF) normalized to 1 above threshold

To normalize to a given total count multiply by the count.

The fitted power-law function normalized to 1 above threshold

xvals: the values at which the fit PDF is to be evaluated alpha: the fitted power thresh: threshold value for the fitting process

To normalize to a given total count multiply by the count.

The integral of the power-law fit above the x-values given (reverse CDF) normalized to 1 above threshold

To normalize to a given total count multiply by the count.

Maximum likelihood fit for the coefficient alpha for a distribution of discrete values p(x) = alpha exp(-alpha*x) above a given threshold. Values below threshold will be discarded. If no threshold is specified, the minimum sample value will be used.

distr: name of distribution, either 'exponential' or 'rayleigh' or 'power'

vals: sequence of real numbers

thresh: threshold to apply before fitting - if thresh=None, use the lowest value

Perform Kolmogorov-Smirnov test of the given set of discrete values above a given threshold for the fitted distribution function ex.: KS_test('exponential', vals, alpha, thresh) If no threshold is specified, the minimum sample value will be used.

Returns the KS test statistic and its p-value: lower p means less probable under the hypothesis of a perfect fit

fitdict

Value:

{'exponential': fit_exponential, 'rayleigh': fit_rayleigh, 'power': fi t_power}

fndict

Value:

{'exponential': expfit, 'rayleigh': rayleighfit, 'power': powerfit}

cum_fndict

Value:

{'exponential': expfit_cum, 'rayleigh': rayleighfit_cum, 'power': powe rfit_cum,}