Open jarbet opened 3 months ago
Thanks for catching that @jarbet !!!
The first issue is a typo and will be fixed in PR #9
The second issue is due to the current class(molecular.data$cna)
would return 2 vectors instead of one: "matrix", "array"
. I will try to fix that in #5
The issue of running create.autoencoder()
is due to the upgrade from Keras v2.x to Keras v3.5, the most current version.
Updating the R package keras
to keras3
will resolve this issue. But to accommodate the update of keras3
will need a few following changes as well:
h5df
to keras
.load_model()
.DESCRIPTION
section and others.
In this example,
csnvs
should besnv
In this example I get an error, could be related to the new matrix checks in #5
I recommend double checking that the rest of the examples work after fixing the above 2 bugs