Numbers of events.
Numbers of trials.
Optionalscore: Sample
Scores of the groups. Default 1, 2, ..., k. A group with a NaN score is left out of the regression
(as R's lm omits it), but its counts stay in the pooled proportion and the weights, as in R. Other scores must
be finite, with at least two distinct values. With no events or only events (all proportions 0, or all 1) the test is not defined (an error, as in R).
Chi-squared test for trend in proportions, with the scores in the options:
propTrendTest(x, n, { score: [0, 1, 2, 4] }).
Chi-squared test for trend in proportions. Equivalent to R's
prop.trend.test(x, n, score). The scores are the third argument or the optionscore(giving both is a RangeError).