Closed mattmoo closed 6 years ago
Talk to the 'broom' guys. -- Sent from Gmail Mobile
Actually afaics broom
already has quantreg
support. So, maybe you mean something else?
WFM:
library(quantreg)
library(huxtable)
tmp <- rq(Petal.Width ~ Sepal.Length, data = iris)
huxreg(tmp, error_format = '{conf.low} / {conf.high}')
## ─────────────────────────────────────────────────
## (1)
## ─────────────────────────
## (Intercept) -3.480
## -4.291 / -2.837
## Sepal.Length 0.800
## 0.656 / 0.970
## ─────────────────────────
## N 150
## logLik -99.889
## AIC 203.777
## ─────────────────────────────────────────────────
## *** p < 0.001; ** p < 0.01; * p < 0.05.
## Column names: names, model1
Thanks Hugh. That looks like almost what I want., but with no statistical testing.
It seems like in quantreg
the stats are passed off to boot.rq
(usually generated using summary(tmp, se="boot")
)
I haven't been able to get tidy
to generate p-values, although maybe that's broom
's issue, as you say.
If summary generates P values then you could try writing a tidy method for summary.quantreg. I think broom is in an awkward place where it can only offer very minimal consistency guarantees. Perhaps that is the inevitable consequence of it being a facade over many different projects. -- Sent from Gmail Mobile
Great package! Is there any chance that support for
quantreg
is on the horizon?