mqr.inference.rate.

confint_1sample

mqr.inference.rate.confint_1sample(count, n, meas=1.0, conf=0.95, bounded='both', method='wald-cc')

Confidence interval for rate count / n / meas.

Parameters:

countint

Number of events.

nint

Number of periods over which events were counted.

measfloat, optional

Extent of one period of observation.

conffloat, optional

Confidence level that determines the width of the interval.

method{‘chi2’, ‘exact’, ‘wald’, ‘wald-cc’}, optional

‘chi2’

Chi2 interval, see [2].

‘exact’

Exact method, recommended for small count. Implements method 9 in [3].

‘wald-cc’

Wald method with continuity correction, recommended for small count. Implements method 5 in [1].

(other)

Everything else is passed to sm..confint_poisson, which supports only two-sided intervals.

Returns:

mqr.inference.confint.ConfidenceInterval

Notes

For a discussion on the benefits and disadvantages of various intervals, including these, see [1] and [3].

References

[1] (1,2)

Patil, V. V., & Kulkarni, H. V. (2012). Comparison of confidence intervals for the Poisson mean: some new aspects. REVSTAT-Statistical Journal, 10(2), 211-22.

[2]

Garwood, F. (1936). Fiducial limits for the Poisson distribution. Biometrika, 28(3/4), 437-442.

[3] (1,2)

Barker, L. (2002). A comparison of nine confidence intervals for a Poisson parameter when the expected number of events is ≤ 5. The American Statistician, 56(2), 85-89.