Is it possible with dplyr or with other R package use any custom cumulative function (i.e. cumulative linear trend, ) for given longitudinal data frame, with next variables:

x - numerical or categorical variable

ID - object ID

t - tenure (1,2,....)

For instance I would like to have linear trend for x[1:2], x[1:3], x[1:4],....for each ID and tenure

in the flexible way (with some dplyr function)

I'm not sure I totally understand your scenario, but are you talking about something akin to window functions?

You might also look at the window functions and functionality provided by the tsibble package

https://tsibble.tidyverts.org/articles/window.html

GIF from Earo Wang's talk on tsibble at rstudio::conf(2019)

Many thanks! Yes, I was thinking about "stretch_dbl"

If your question's been answered (even by you!), would you mind choosing a solution? It helps other people see which questions still need help, or find solutions if they have similar problems. Here’s how to do it:

Hi. I can not see where I can close my question

Here's a GIF of how you do it (I selected and then de-selected so that you can do it yourself)

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.