... at least for dummies :)

There is already a invert.x option plot() for ZCurvesDataframe and ZGroupCurvesDataframe, but this only affects the labels on the x-axis. Would be useful if the plotting order could be changed.

So far more-or-less necessary packages are all in depends in DESCRIPTION file. See if some of these could be moved to imports or suggests.

bat-files created by create_zproject() include trailing whitespaces, remove these.

Simply do:

run_bat([email protected])

Transfer most of the old Depends packages to @import or @importFrom.

If there are many features / groups to plot in performance curves, the graphical resolution of the plot may not be good. Plot generic should inspect how many lines need to be plotted and adjust the linetype/color mappings accordingly.

Feturenames are only assigned if set explicitly. They should be set even is using the default ones.

E.g. raster path in spp file

C:Data\foo\bar\some_spam.img

becomes name foo bar some_spam

Group IDs vector can be changed:

groups(variant) <- something

but trying to add new names mapping does not work as the number of categories is still being checked against the original groups-file.

Using argument cols while groups=TRUE in curves() does not work. Either warn user about this or implement correct group subsetting.

get_variant(x, index) where x is a Zproject object and index is either a character or numeric index (name or a position of a variant). index should be checked for validity.

Both plot_curves and plot_grp_curves should have only 1 plotting function. Preprocessing the data (read in and parse) should be done separately (still within 1 function?).

Simple functionality for writing dat (ini) files needed.

It would be good to have this option, as e.g. the project maker makes projects so that the bat file is written under the variant folder.

Method should save all the associated (Zonation input) files on the file system.

create_zproject() tries to do too much now: it can create a new Zonation project as well load existing. Create a new function load_zproject() for cleaner and more intuitive API.

performance(x, top.fraction)

x = Zvariant object
top.fraction = Top fraction (in percentage!) of the landscape for which the performance level is returned. E.g. top.fraction=15 would equal to rank >= 0.85. Can be a single value or a vector of values.

value = A data frame with given percentages as columns and feature as rows.

Now groups data and spp data are stored separately, these should be merged internally

When Zproject is being created based on existing project, a lot of files are being read in. Implement debug logging to track which files are read in in case something goes wrong.

Similar to groupnames<- / groupnames

Right now parse_bat() will return a command string, returning a data.frame with the command elements would be more versatile

Does not find zig(3or4).exe in windows, even if the path variable is correct.

This would probably call for example data. Could the tutorial be used?

regular rasters are just fine

Create functions get_dat_param() and set_dat_param() methods for class Zvariant. These work with @dat.data slot.

performance only returns pr_lost and feature levels or group means. Overall average would be useful too.

Normal conditional message statements should be enough.

My simple Z run didn't have groups, and thus 'results' considered my results to be empty, although they were not.

Use underscore "_" instead of dot or camelCase notation.

At least when groups have not been used initially.

This is a simpler version of create_zproject() which creates just a single variant.

Create_zproject would benefit from a feature where user could define different paths to be used in .bat files. Now root.path is used automatically which is awkward if one wants to use relative paths.

Feature names works ok for variants but get lost when analyzing results (e.g. the plots use only f1, f2, f3, etc. and featurenames(results) return this:
[1] "f1" "f2" "f3" "f4"