We would like these to be equivalent, but they are not. Presumably, %>% is forcing evaluation, so repeat_n() is simply repeating the already evaluated code. Not sure whether there is a workaround for this.
> lm(length ~ width, data = resample(KidsFeet)) %>% repeat_n(3)
term estimate std.error statistic p.value
1 (Intercept) 6.093808 4.059640 1.501071 1.418236e-01
2 width 2.064542 0.446912 4.619572 4.539183e-05
3 (Intercept) 6.093808 4.059640 1.501071 1.418236e-01
4 width 2.064542 0.446912 4.619572 4.539183e-05
5 (Intercept) 6.093808 4.059640 1.501071 1.418236e-01
6 width 2.064542 0.446912 4.619572 4.539183e-05
> repeat_n(lm(length ~ width, data = resample(KidsFeet)), 3)
term estimate std.error statistic p.value
1 (Intercept) 6.468877 2.8536714 2.266861 2.933423e-02
2 width 2.030487 0.3133226 6.480500 1.408364e-07
3 (Intercept) 3.081941 2.7112242 1.136734 2.629583e-01
4 width 2.422475 0.3009253 8.050087 1.188295e-09
5 (Intercept) 13.620874 3.7710251 3.611982 8.969603e-04
6 width 1.236907 0.4168456 2.967301 5.240899e-03
We would like these to be equivalent, but they are not. Presumably,
%>%is forcing evaluation, sorepeat_n()is simply repeating the already evaluated code. Not sure whether there is a workaround for this.