I try to install songbird in my conda environment for qiime:
(qiime2-2020.2) >>pip install songbird
I get the following error:
ERROR: Could not find a version that satisfies the requirement tensorflow<2,>=1.15 (from songbird) (from versions: 2.2.0rc3, 2.2.0rc4, 2.2.0, 2.3.0rc0, 2.3.0rc1, 2.3.0rc2)
ERROR: No matching distribution found for tensorflow<2,>=1.15 (from songbird)
Any idea about how to work around that issue?
Have you tried the conda install? pip and conda are known to not play together well.
I have just fixed the issue. I have followed these instructions from https://github.com/knightlab-analyses/reference-frames
conda create -n songbird_env python=3.6 numpy=1.15.4 scikit-bio=0.5.5 seaborn pandas=0.23.4 -c conda-forge
source activate songbird_env
conda install tensorflow=1.10 tqdm nomkl
conda install biom-format h5py -c conda-forge
conda install jupyter notebook
conda install songbird -c conda-forge
Then in qiime2-2020.6 environment, I have simply run
pip install songbird
qiime dev refresh-cache
to refresh qiime.
Not sure I totally understand what the issue was…
ok … the instructions from the reference-frames repo is to install the standalone version – that won’t install songbird with qiime2.
I’m still a bit confused why the
qiime dev refresh-cache command worked, but glad that you found a solution!
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.