Following {tidyselect} (dep for {dplyr}), .data$x is now "x" in most
{dplyr} selections verbs.

This is explained in :
https://community.rstudio.com/t/use-of-data-in-tidyselect-expressions-is-now-deprecated/150092/3

Removed old files in vignettes folder
Run the vignettes with check to remove file names warnings

Merge the summary branch and added the functions for ipm and species.
Forest need to be done. Coverage is done.

Added delay to the main vignette and finished the Harvesting vignette.
Still missing Even scenario harvest.

Some dev files have been removed with the old multisp_deter_sim matrix.

Still in working progress, need to end Uneven and single species Even
harvest.

`sim_deter_forest()` output is modified and now direclty return de
`tree_format` tibble.

Vignettes have evolved to match this usage.

Delayed IPM caused NaN during division with survival due to mesh equal
to 0.

Also delayed mesh size are not counted in quadratic mean diameter.

Multi species simulations should works but don't, will fix tomorrow

Single simulation works but disturbance duration is not perfect.

Modified the IPM class and get_step_IPM function so that we can
desactivate survival during simulation.

Now we can compute Mu values and store them in correct matrix.

A new function build the IPM matrix corresponding to the BA and climate,
ready to be used at each step of the simulation with some more tweaks.

This function takes 15ms so it's a bit long, and I try to bring time
down.

Computing mu values for a species is now down.