algorithms.stopping_rules

Module: algorithms.stopping_rules

Stopping rules used in sequential FDR control.

See `http://arxiv.org/abs/1309.5352`_

Functions

selectinf.algorithms.stopping_rules.forward_stop(pvalues, alpha)

Compute the number of rejections using forward stop of `http://arxiv.org/abs/1309.5352`_

>>> forward_stop(np.array([0.5,0.6,0.7,0.8,0.9]), 0.05)
0
>>> forward_stop(np.array([0.001, 0.002, 0.0015, 0.0013, 0.05, 0.6]), 0.05)
5

In R:

> forwardstop(c(0.5,0.6,0.7,0.8,0.9), 0.05) [1] 0 > forwardstop(c(0.001, 0.002, 0.0015, 0.0013, 0.05, 0.6), 0.05) [1] 5 >

Parameters

pvalues : np.float

alpha : float

Returns

num_rejections : int

selectinf.algorithms.stopping_rules.simple_stop(pvalues, alpha)

Compute the number of rejections using simple stop, the first time a p-value is above alpha.

Parameters

pvalues : np.float

alpha : float

Returns

num_rejections : int

selectinf.algorithms.stopping_rules.strong_stop(pvalues, alpha)

Compute the number of rejections using strong stop of `http://arxiv.org/abs/1309.5352`_

>>> strong_stop(np.array([0.5,0.6,0.7,0.8,0.9]), 0.05)
0
>>> strong_stop(np.array([0.001, 0.002, 0.0015, 0.0013, 0.05, 0.6]), 0.05)
3

In R:

> strongstop(c(0.001, 0.002, 0.0015, 0.0013, 0.05, 0.6), 0.05) [1] 3 > strongstop(c(0.5,0.6,0.7,0.8,0.9), 0.05) [1] 0

Parameters

pvalues : np.float

alpha : float

Returns

num_rejections : int