I the userguide n chapter 9.1 it is said to use the following command however it gives a weird error in the PlantGL 3D Viewer:

l.plot(tree)

Are you sure that it should be simply the following?

plot(tree)

When I do that the plot is almost correct (except that the angles are assumed to be always 45 degrees)

Add a method to dump and load the MTG into/from a file.
The method needs to check if all the properties can be pickled.

The function mtg2lpy() is an extremely useful function

Unfortunately there is not much documentation yet

I am trying to use it but I don't understand which argument I should use.

I see there are 2 arguments:
the first should be the actual .mtg (I called it g and I generated with g = read_mtg_file('test.mtg'))
the second argument is supposed to be a 'lsystem' but I really don't understand what should it be since I have no lsystem yet (that's what I am trying generate with mtg2lpy!)

Thank you very much!

reading a valid MTG fails under Python 3 because io.py uses bad_right as a boolean. It needs to check something like len(list(bad_right) > 0 instead . See line 1041 and 1042

Hello,
apparently in the later versions of python StopIteration exception raises a runtime error. Here there is more on that:
https://peps.python.org/pep-0479/

Therefore we have to change the insert_scale method code, in particular line 1099

sup_components = dict((v, next(g.component_roots_iter(v))) for v in sup_vertices)

raises a runtime error and stops execution. I replaced it with

sup_components = dict()
for v in sup_vertices:
iteratorr = g.component_roots_iter(v)
while True:
try:
sup_components[v] = next(iteratorr)
except StopIteration:
break

And now it seems to work fine.
Can you verify that this is correct ?

After removing a vertex with remove_vertex, properties are not updated.
further using _remove_vertex properties does the job.
Probably due to wrong call in tree.py (remove vertex does not return vid), line 535,536:

vid = super(PropertyTree, self).remove_vertex(vid, reparent_child=reparent_child) self._remove_vertex_properties(vid)

guessing that :

super(PropertyTree, self).remove_vertex(vid, reparent_child=reparent_child) self._remove_vertex_properties(vid)

should be okay

Some MTG may have missing vertices at the lower scale.

ENTITY-CODE						
/P									
/A1											2016
^/S1
	+B1										2017
		+C1									2018

C1 has no parents

g.roots(scale=2) return [2, 4, 5]

Add test to conda.
Then at each modifications, all the tests will be run

Based on plantgl I guessed that the install instruction was

conda install openalea.mtg -c fredboudon -c conda-forge

but could that be added to the README?

Querying a mtg with [] always returns a result, even if the vid is not in the mtg, which is a bit confusing.

Create an MTG from a DataFrame.

When building an MTG from file, there are a set of warning that are displayed.
It would be good to add a silent or verbose mode to allow the user to manage the log information

• Add a verbose mode to the MTG constructor
• Remove debug information with dynamic MTG

I tried to install openalea.mtg via conda on Windows.
conda install -c openalea3 openalea.mtg

A package conflict message appears see :

Not having previously installed python I installed python and tried again
conda install -c conda-forge python
conda install -c openalea3 openalea.mtg

I had a conflict on my python 3.10.4 version
so I downgraded to 3.10.0
conda install -c conda-forge python=3.10.0
conda install -c openalea3 openalea.mtg
it still didn't work

same for python=3.9.12
conda install -c conda-forge python=3.9
conda install -c openalea3 openalea.mtg

Use modernize to port the coed on both Python2 and Python 3

The MTG reader fail strongly when some column names are not defined in the header (FEATURE).
The reader seems to shift the columns by non considering the unknow property names.

Please raise an error when this is the case.
Then the header can be fixed easily.

While using the above function on an mtg which is the output of hydroshoot, some values appear displaced. Notably it seems that the presence of an array within the properties, i.e.
'TopPositionPoints': array([[ 5.39, -7.29, 48.87],
[ 5.48, -7.57, 47.6 ],
[ 4.61, -7.44, 46.97],
[ 4.06, -6.92, 45.6 ],
[ 3.63, -6.07, 46.86]])

confuses the function and causes a displacement of a number of values. It happens in all nodes which represent leaves, where a similar array is present in the properties.