search

rrwick / Deepbinner

a signal-level demultiplexer for Oxford Nanopore reads
GNU General Public License v3.0
124 stars 23 forks source link

What am I doing wrong? #14

Open xgnusr opened 5 years ago

xgnusr commented 5 years ago

Hello, after installing Deepbinner:

....... usr@pandora:/test/Deepbinner/models$ deepbinner classify -s SQK-RBK004_read_starts input Using TensorFlow backend. Loading SQK-RBK004_read_starts... done Error: input is neither a file nor a directory .......

Another try:

usr@pandora:/albacore/gridion/basecalled/01010102/GA10000$ deepbinner classify --rapid /albacore/gridion/basecalled/01010102/GA10000/reads/0/ > classifications Using TensorFlow backend. 2018-11-09 15:13:00.956983: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA Loading /usr/local/lib/python3.6/site-packages/deepbinner/models/SQK-RBK004_read_starts... Traceback (most recent call last): File "/usr/local/bin/deepbinner", line 11, in sys.exit(main()) File "/usr/local/lib/python3.6/site-packages/deepbinner/deepbinner.py", line 60, in main classify(args) File "/usr/local/lib/python3.6/site-packages/deepbinner/classify.py", line 36, in classify load_and_check_models(args.start_model, args.end_model, args.scan_size) File "/usr/local/lib/python3.6/site-packages/deepbinner/classify.py", line 66, in load_and_check_models out_dest=out_dest) File "/usr/local/lib/python3.6/site-packages/deepbinner/classify.py", line 90, in load_trained_model model = load_model(model_file) File "/usr/local/lib/python3.6/site-packages/keras/engine/saving.py", line 417, in load_model f = h5dict(filepath, 'r') File "/usr/local/lib/python3.6/site-packages/keras/utils/io_utils.py", line 197, in init 'Received: {}.'.format(type(path))) TypeError: Required Group, str or dict. Received: <class 'pathlib.PosixPath'>.

..................................

Thank's in advance ......

giesselmann commented 5 years ago

same error here.

giesselmann commented 5 years ago

it happens, if you use the --rapid or --native flags, if you specify the models -s / -e manually, it works.

lowandrew commented 5 years ago

I'm having the same issue, although the above workaround works for me too.

nicoledel3 commented 5 years ago

What did you specify after the -s / -e flags? I'm trying to use the --rapid read start model, but I got the same error. Thank you!

nicoledel3 commented 5 years ago

What did you specify after the -s / -e flags? I'm trying to use the --rapid read start model, but I got the same error. Thank you! To clarify I specified -s SQK-RBK004 and got that SQK-RBK004 didn't exist.

lowandrew commented 5 years ago

I downloaded the rapid read model from this repo and then gave the path to that. Alternatively, the pull request that's waiting to be merged fixes things so that --rapid and --native can be used. You'd just have to clone that and then pip install -e . from the base of the repository.

ttubb commented 5 years ago

For anyone finding this thread: You can make a small change in deepbinner.py by yourself, explained here: https://github.com/eclarke/Deepbinner/commit/00e9ff93a0ffa21f583e258cd071fec00ae4f0a8 .