dependency conflicts during installation
Closed this issue · 3 comments
I want to run FRETboard locally but am running into dependency conflicts as well as a failed build for quast.
Failed to build quast
DEPRECATION: Could not build wheels for quast which do not use PEP 517. pip will fall back to legacy 'setup.py install' for these. pip 21.0 will remove support for this functionality. A possible replacement is to fix the wheel build issue reported above. You can find discussion regarding this at https://github.com/pypa/pip/issues/8368.
Installing collected packages: six, smmap, gitdb, gitpython, pyparsing, docutils, amply, pulp, psutil, ratelimiter, configargparse, wrapt, ipython-genutils, traitlets, jupyter-core, attrs, pyrsistent, zipp, importlib-metadata, jsonschema, nbformat, pyYAML, appdirs, urllib3, chardet, idna, requests, toposort, datrie, snakemake, mappy, numpy, biopython, ete3, retrying, plotly, pytz, python-dateutil, pandas, pysam, nanoget, pillow, kiwisolver, cycler, matplotlib, Python-Deprecated, nanomath, pyarrow, scipy, seaborn, pauvre, NanoPlot, ptyprocess, pexpect, joblib, simplejson, quast, poreTally
Running setup.py install for quast ... done
ERROR: After October 2020 you may experience errors when installing or updating packages. This is because pip will change the way that it resolves dependency conflicts.
We recommend you use --use-feature=2020-resolver to test your packages with the new resolver before it becomes the default.
thinc 7.0.4 requires tqdm<5.0.0,>=4.10.0, which is not installed.
tensorboard 1.13.1 requires werkzeug>=0.11.15, which is not installed.
scikit-image 0.14.2 requires cloudpickle>=0.2.1, which is not installed.
scikit-image 0.14.2 requires dask[array]>=1.0.0, which is not installed.
scikit-image 0.14.2 requires networkx>=1.8, which is not installed.
scikit-image 0.14.2 requires PyWavelets>=0.4.0, which is not installed.
mindsdb 1.0.7 requires cython>=0.29.2, which is not installed.
mindsdb 1.0.7 requires xlrd>=1.0.0, which is not installed.
ludwig 0.1.1 requires Cython>=0.25, which is not installed.
ludwig 0.1.1 requires h5py>=2.6, which is not installed.
ludwig 0.1.1 requires tqdm, which is not installed.
keras-applications 1.0.7 requires h5py, which is not installed.
spacy 2.1.3 requires jsonschema<3.0.0,>=2.6.0, but you'll have jsonschema 3.2.0 which is incompatible.
Hi mayeshh!
Thanks for your interest in FRETboard. I see the issue; I accidentally gave install instructions for the previous tool I wrote, which was something else entirely. The correct install command should be:
pip install git+https://github.com/cvdelannoy/FRETboard.git
Force of habit I guess ^^' sorry for the confusion, the readme is also updated now!
Btw, if you're still experiencing issues with dependency conflicts I can recommend using a clean virtualenv or conda environment. Let me know if it works out now!
Thanks for your quick response!! I was wondering about the url :)
I did clean reinstallation in a python 3.8 env, and there were compatibility issues. It seems that the setup works best with python 3.7. After downgrading my version of python, there is still the following error:
ERROR: After October 2020 you may experience errors when installing or updating packages. This is because pip will change the way that it resolves dependency conflicts.
We recommend you use --use-feature=2020-resolver to test your packages with the new resolver before it becomes the default.
thinc 7.0.4 requires tqdm<5.0.0,>=4.10.0, which is not installed.
tensorboard 1.13.1 requires werkzeug>=0.11.15, which is not installed.
spacy 2.1.3 requires jsonschema<3.0.0,>=2.6.0, which is not installed.
spacy 2.1.3 requires requests<3.0.0,>=2.13.0, which is not installed.
scikit-image 0.14.2 requires cloudpickle>=0.2.1, which is not installed.
scikit-image 0.14.2 requires dask[array]>=1.0.0, which is not installed.
scikit-image 0.14.2 requires PyWavelets>=0.4.0, which is not installed.
mindsdb 1.0.7 requires attrs>=18.2.0, which is not installed.
mindsdb 1.0.7 requires chardet>=3.0.4, which is not installed.
mindsdb 1.0.7 requires requests>=2.20.0, which is not installed.
mindsdb 1.0.7 requires urllib3>=1.23, which is not installed.
mindsdb 1.0.7 requires xlrd>=1.0.0, which is not installed.
ludwig 0.1.1 requires tqdm, which is not installed.
Looking through setup.py I installed the correct version:
conda install scikit-learn==0.21.2
This worked.
I will close this issue.
Odd, I haven't encountered any issues with python 3.8 or dependency issues with sklearn...thanks for reporting it, I'll dig a bit further some day to see if I can fix it.