A climate_network seems to leave the initial value of non_localuntouched regardless of changing it with set_non_local(). Here is an example showing that after toggling the attribute the triggered function is the same:
net = climate.TsonisClimateNetwork(data, threshold=0, winter_only=False, non_local=True)
Generating a Tsonis climate network...
Calculating correlation matrix at zero lag from anomaly values...
Extracting network adjacency matrix removing local connections...
Calculating angular great circle distance using Weave...
Extracting network adjacency matrix by thresholding...
Setting area weights according to type surface...
print(net)
Tsonis climate network:
...
Local connections filtered out: True
Use only data points from winter months: False
net.set_non_local(False)
Extracting network adjacency matrix removing local connections...
Calculating angular great circle distance using Weave...
Extracting network adjacency matrix by thresholding...
Setting area weights according to type surface...
print(net)
Tsonis climate network:
...
Local connections filtered out: True
Use only data points from winter months: False
The same behaviour is shown for networks initially set to non_local=False, just the other way around.
Using the parameter explicily (net.set_non_local(non_local=True)) does not help.
A climate_network seems to leave the initial value of
non_local
untouched regardless of changing it withset_non_local()
. Here is an example showing that after toggling the attribute the triggered function is the same:The same behaviour is shown for networks initially set to
non_local=False
, just the other way around. Using the parameter explicily (net.set_non_local(non_local=True)
) does not help.