Issues in collect samples function about deer HOT 5 OPEN

@AdritaBarua By looking at the number shown in the status bar, I guess you are trying to generate samples where the target graph is undirected and the source graph is directed. This will bring up this problem since the "find_path_between_pair" tries to find a path starting from the "first_node" and ending in "second_node". When the target graph is undirected, there is no direction between the two nodes of an edge, and if the source graph is directed, the "second_node" may not be in the neighbors collected by graph.neighbors(first_node).

So to avoid this error, you may try target-source graphs as "directed-directed", "directed-undirected" or "undirected-undirected" instead of "undirected-directed". If you want to implement the "undirected-directed" collection, you may modify the following code

first_neighbors = set(graph.neighbors(first_node))
first_neighbors.remove(second_node)
second_neighbors = set(graph.neighbors(second_node) if not graph.is_directed() else graph.predecessors(second_node))
second_neighbors.remove(first_node)

into

first_neighbors = set(graph.neighbors(first_node))
if second_node in first_neighbors:
    first_neighbors.remove(second_node)
second_neighbors = set(graph.neighbors(second_node) if not graph.is_directed() else graph.predecessors(second_node))
if first_node in second_neighbors:
    second_neighbors.remove(first_node)

from deer.

Thanks for your response, we tried all the possibilities, and even with digraph-digraph, still we are getting the same error. and the previous run was for undirected-undirected. Could you please let us know where we have a problem, that we are getting this error?

from deer.

@AdritaBarua To run for undirected-undirected, you may place the generated "graph.pickle" under the "extract_wiki" folder and then run python extract_wiki.py collect_sample false false 0.75. This command works on my computer.

from deer.

@ZhuKerui We generated the "graph.pickle", but the error for running collect_sample as you can see in the screenshot is for the " Key error: Henry II, Duke of Bavaria"

from deer.

@sanazskn Hi, could you please show me the complete error message and the command you use to run the script? Thanks.

from deer.