def test_train(datapath, jobid, wpc, lr_opt, batch_size, epochs, l1_value, problem_type, ):
test1 = os.system(
'cd .. && python GenNet.py train {datapath} {jobid} -problem_type'
' {problem_type} -wpc {wpc} -lr {lr} -bs {bs} -epochs {epochs} -L1 {L1}'.format(
datapath=datapath, jobid=jobid, problem_type=problem_type, wpc=wpc, lr=lr_opt, bs=batch_size, epochs=epochs,
L1=l1_value))
assert test1 == 0
folder, resultpath = get_paths(jobid=jobid)
test2 = os.path.exists(resultpath + '/bestweights_job.h5')
assert test2
def plot(args):
folder, resultpath = get_paths(args.ID)
importance_csv = pd.read_csv(resultpath + "/connection_weights.csv",
index_col=0)
print(resultpath)
layer = args.layer_n
if args.type == "layer_weight":
plot_layer_weight(resultpath,
importance_csv,
layer=layer,
num_annotated=10)
elif args.type == "sunburst":
sunburst_plot(resultpath=resultpath, importance_csv=importance_csv)
elif args.type == "raw_importance":
manhattan_importance(resultpath=resultpath,
importance_csv=importance_csv)
else:
print("invalid type:", args.type)
exit()