Diff in means/props/medians

[mine-cetinkaya-rundel]

Currently I believe the subtraction is being done in the alphabetical order of the levels if the categorical variable is character, and in the levels order if variable is factor. I think we need an order argument for calculate for "diff in ..."situations. This is how I had solved this problem in the inference function. Happy to do a PR but wanted to see (1) am I interpreting current status correctly and (2) is everyone on board with a new order argument that will be something like order = c("level1" - "level2") which translates to mean(y[x == "level1"]) - mean(y[x == "level2"]).

[ismayc]

I’m on board with this and it makes sense to me. Can you make a pull request to the dev branch? Or would you like this implemented in version 0.0.1 on master?

[mine-cetinkaya-rundel]

I would like it implemented. I think the feature is mandatory, I’m not sure if there is a better implementation.

On Sat, Nov 4, 2017 at 1:12 PM Chester Ismay ***@***.***> wrote: I’m on board with this and it makes sense to me. Can you make a pull request to the dev branch? Or would you like this implemented in version 0.0.1 on master? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_andrewpbray_infer_issues_53-23issuecomment-2D341913671&d=DwMFaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=VMaFnDxOO7DFh_jmcYrq6Oi2_V_d1YoR6dsQ-kCSJjw&m=DL6BR_MGTpnXgaCpX4SVh30XaoBlpqq0C4gORPDhdlw&s=SNVyuk5L8dEXC1FKcLpvdY45SItW1XjWjIV34PKczJo&e=>, or mute the thread <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AFsHUaagGB4g8sR-2DtrwP1oO-5F9h9jEzadks5szJsUgaJpZM4QSESZ&d=DwMFaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=VMaFnDxOO7DFh_jmcYrq6Oi2_V_d1YoR6dsQ-kCSJjw&m=DL6BR_MGTpnXgaCpX4SVh30XaoBlpqq0C4gORPDhdlw&s=Rp7Ca63kU6iQ6HcmFU4YvtrXcf3zZnSAFQP9DCaUrKA&e=> .

-- Mine Çetinkaya-Rundel, Ph.D. Director of Undergraduate Studies Associate Professor of the Practice Duke University, Department of Statistical Science <mine@stat.duke.edu> mine@stat.duke.edu <https://stat.duke.edu/~mc301/> https://stat.duke.edu/~mc301 <https://github.com/mine-cetinkaya-rundel> <https://twitter.com/minebocek> <http://citizen-statistician.org/>

[ismayc]

@andrewpbray: Mine will be sending a pull request to master to implement this. Maybe update the package version to 0.0.2 as well with 0.1.0 having the theoretical distribution stuff on the dev branch?

[mine-cetinkaya-rundel]

For the record, I was wrong earlier, subtraction is done in reverse alphabetical order since calculate uses diff which works as follows:

> x <- c(1,3)
> diff(x)
[1] 2

[andrewpbray]

@mine-cetinkaya-rundel what's the status of this? The addition of the order argument should work with the new formulation of success in specify.

[mine-cetinkaya-rundel]

I'm working on it -- that's where the other question came about.

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.