matching API#

Simple matching functionalities for RSML MTG trees

openalea.rsml.matching.match_plants(t1, t2, max_distance=None)[source]#

Find a 1-to-1 matching between plants in t1 & t2

The matching is done to minimize distance between plants “seed”: the mean of the first node of all primary axes.

This function returns:

  • the set of matched (t1-plant-id,t2-plant-id,distance)

  • the set of unmatch t1 plant id

  • the set of unmatch t2 plant id

The matching is done usinf one_to_one_match

openalea.rsml.matching.match_roots(t1, t2, plant_match, max_distance=None)[source]#

Iteratively match root axes following the trees topology

This function iteratively match root axes from 1st to last (topological) order. The matching is selected using one_to_one_match using hausdorff distance.

Inputs:

t1: 1st rsml-mtg t2: 2nd rsml-mtg plant_match: match between plants in t1 and t2

Outputs:

  • the set of root matches as (t1_root_id, t2_root_id, hausdorff-distance)

  • the set of unmatched root in t1

  • the set of unmatched root in t2

Note

Children of unmatched elements are not listed in unmatched sets

openalea.rsml.matching.one_to_one_match(distance, max_distance=None)[source]#

select minimal 1-to-1 matching in distance matrix, iteratively

Iteratively select the match (i,j) with minimum value in distance but only if both i and j were not matched at a previous step.

If max_distance is not None, don’t match pairs with a distance value higher than given max_distance

Return:

  • the set match {(i0,j0,d0),(i1,j1,d1),…} where i*,j*,d* are the 1st, 2nd indices, and their distance

  • the set of unmatch i

  • the set of unmatch j

example:

d = np.random.randint(0,10,5,4)
m,i,j = direct_match(d)
print '

‘.join(str(ij)+’ matched with d=’+str(di) for ij,di in zip(m,d[zip(*m)]))

Download the source file ../../src/openalea/rsml/matching.py.