NEWS

treePlotArea 3.1.0.9000

Internal optimization: Now passing down argument counting_factor from get_correction_factors() to get_correction_factor() to check_tree() to get_boundary_radius() and to all calls to get_r_max().

Incorporate gitlab issues #1, #2 and partly #3: make plot_tree_plot_area
- use asp = TRUE,
- have an argument counting_factor
- return a list with the plot's boundaries and trees alongside the correction factors instead of a vector of correction factors.

Added argument is_ti_round defaulting to TRUE to get_corrections_factors() to allow for rounding the tree's boundary circle's radius to the unit (i.e. [cm]) as done by Thuenen Institute?

As requested by Gerald Kaendler, 'plot_tree_plot_area()' now plots all trees of a corner, if no tree number is given.
To allow for using the field data, I have added missing dbh to the internal checks, albeit it being reported by validate_data(). Just in case I should forget to validate data first.

Fixed validate_data() for boundaries (we accidentally deleted the flexing points).
Now passing argument counting_factor from get_corrections_factors() down.
It is now possible to disable checking the angle counts in get_corrections_factors() via argument skip_check = TRUE. This enables us to check for concentric circles otherwise marked as invalid an given correction factors of 0.

We now check that trees include the corner's center in their plot area and otherwise assign a correction factor of 0. Such trees occur when the diameters where not measured at breast height and got corrected. Otherwise, bwi2022de would have complained about them being invalid sample trees.
Streamlined internal status codes.

Exported function check_boundaries(). It looks for invalid boundaries, one of which occurs in the federal database of the 2012 survey.
select_valid_angle_count_trees() now removes trees with a diameter at breast height greater than zero and a distance of 0, for these tree should not be there. Gerald/grenzkreis gives a warning, but assigns a correction factor >= 1. Probably, these trees get filter out somewhere down the food chain. We assign a correction factor of 0 instead of only issue a warning.

We are checking for boundaries that happen to run through a tract's corner. It's not even clear how we would define stand if such a boundary would be valid. Despite that we by default keep and not delete these boundaries since they were dealt with by grenzkreis. So we do that, too. You may get rid of them using get_boundary_polygons(..., clean_data = TRUE) explicitly.
get_corrections_factors() now adds status codes and info to the output.
Added testing suite for internal function get_corrections_factor, which is the main work horse in the package (to which get_corrections_factors() is just a wrapper looping over the trees.
Added verbosity to internal function secant_intersection() to not throw warnings all over the place.
Added internal function tree_and_boundary() for manually double checking on trees and boundaries. It's not documented but you may use it to check on constellation where you have doubts.
Added validate_data() to validate data to conform to the standards of the federal database.
get_corrections_factors() now returns NA instead of stopping on any error.

Many minor data, docs and testing stuff done. Package is now quite well tested, but documentation lacks. And we need to pass column names.