Compute t-statistic (TScore)

class cerebunit.statistics.stat_scores.tScore.TScore(*args, **kwargs)

Compute t-statistic as the standardized statistic as

Definitions	Interpretation
sample_mean, \(\bar{x}\)	observation[“mean”]
null_value, \(\mu_0\)	model prediction
standard_error, se	observation[“standard_error”]
t-statistic, t	t = \(\frac{\bar{x} - \mu_0}{se}\)

Note: se = \(\frac{s}{\sqrt{n}}\), where n is the sample size and s is the standard deviation.

Use Case

x = TScore.compute( observation, prediction )
score = TScore(x)

Note: As part of the SciUnit framework this custom TScore should have the following methods,

• compute() (class method)
• sort_key() (property)
• __str__()

classmethod compute(observation, prediction)

Note:

• observation (sample) is in dictionary form with keys mean and
• standard_error whose value has magnitude and python quantity
• the populations parameter is the predicted value