search

marioangst / motifr

Motif analysis in multi-level networks
https://marioangst.github.io/motifr/
Other
13 stars 2 forks source link

convert character level attributes to factors #27

Closed marioangst closed 4 years ago

marioangst commented 4 years ago

If motif_summary is given a lvl_attr that is just characters, it will not pass the needed integers to sma.

Solution: enforce conversion to numeric via factors (and throw warning to user)

marioangst commented 4 years ago

@tseppelt I think this is probably better dealt with at the Python side as providing character attributes to all other functions seems to work fine. minim reprex:

# set a character vertex attribute based on sesType
ml_net <- network::set.vertex.attribute(ml_net, "lvl", as.character(network::get.vertex.attribute(ml_net,"sesType")))
# check it worked
network::get.vertex.attribute(ml_net,"lvl")

# does not work (setting lvl_attr to "lvl")
motif_summary(ml_net, lvl_attr = "lvl")
#gives:
# Fehler in py_call_impl(callable, dots$args, dots$keywords) : 
# TypeError: 'float' object cannot be interpreted as an integer 

# does work:
exemplify_motif(ml_net, motif = "1,2[I.C]", directed = FALSE, lvl_attr = "lvl")
show_motif("1,2[I.C]", net = ml_net, directed = FALSE, label = TRUE, lvl_attr = "lvl")
tseppelt commented 4 years ago

The reason for this was indeed a problem on the Python side. Now resolved in this commit. Wait for the next version of sma or update manually.

tseppelt commented 4 years ago

Version 2.1.1 of sma which resolves the issue is online on PyPI. As of d359731, there is a test for the workaround.