Hi,
I previosly submitted code and had previous issues I found.
While benchmarking the vae that estimates cell type specific expression, I found a new error.
I am using the up-to-date version of keras, which is the new default.
Keras3 changed their layer module from merge to merging. There you find the code for concatenate
https://github.com/keras-team/keras/tree/master/keras/src/layers/merging
The new names in keras3 causes an error in dissect because you are importing
from keras.layers.merge import concatenate as concat
That causes the error
dissect_expr.py", line 26, in <module>
from keras.layers.merge import concatenate as concat
ModuleNotFoundError: No module named 'keras.layers.merge'
The proportion estimation works but the second model doesnt work. I will change code and test whether it works and will
submit a pull request that supports the new keras API later.
Best regards,
Sergej Ruff4
Hi,
I previosly submitted code and had previous issues I found.
While benchmarking the vae that estimates cell type specific expression, I found a new error.
I am using the up-to-date version of keras, which is the new default.
Keras3 changed their layer module from merge to merging. There you find the code for concatenate
https://github.com/keras-team/keras/tree/master/keras/src/layers/merging
The new names in keras3 causes an error in dissect because you are importing
from keras.layers.merge import concatenate as concatThat causes the error
The proportion estimation works but the second model doesnt work. I will change code and test whether it works and will
submit a pull request that supports the new keras API later.
Best regards,
Sergej Ruff4