I want to install Faiss-GPU on Lambda Stack by conda install -c pytorch faiss-gpu but there is no conda installed.
I tried the solution mentioned here: Installing faiss on Google Colaboratory with this version of faiss https://anaconda.org/pytorch/faiss-gpu/1.6.0/download/linux-64/faiss-gpu-1.6.0-py36h1a5d453_0.tar.bz2 but is says:
ImportError: libmkl_intel_lp64.so: cannot open shared object file: No such file or directory
ModuleNotFoundError: No module named '_swigfaiss_avx2'
ImportError: libmkl_intel_lp64.so: cannot open shared object file: No such file or directory
ModuleNotFoundError: No module named '_swigfaiss'
What is the correct way to install Faiss on Lambda Stack?
if you are colab just try the following:
!sudo apt-get install libomp-dev
Then it works.
Related
I installed giotto-tda library with a command: %pip install --user pyproj giotto-tda
and then tried to use: from gtda.time_series import TakensEmbedding
but I am still getting an error: "No module named 'gtda'.
Did you have a similar problem?
I have one error: ModuleNotFoundError: No module named '_sysconfigdata_x86_64_conda_linux_gnu'
Using comand: pip3 install --editable ./BindingSitesFromFragments
I had this problem after installing anaconda (w/ python3.8.8) on Ubuntu 20.04 on an aws ec2 instance.
Here is my error message:
"ModuleNotFoundError: No module named '_sysconfigdata_x86_64_conda_linux_gnu'"
After a search
sudo find / -name _sysconfigdata_x86_64*`
and found the file
/home/ubuntu/anaconda3/lib/python3.8/_sysconfigdata_x86_64_conda_cos6_linux_gnu.py
Copying this file to the name of the missing file solved my problem
cd ~/anaconda3/lib/python3.8
cp _sysconfigdata_x86_64_conda_cos6_linux_gnu.py _sysconfigdata_x86_64_conda_linux_gnu.py
works however, it will throw an another error, Downgrade the pandas to 0.20.3 or 1.4.1 worked for me
#--------------------------------
from pandas.util._exceptions import find_stack_level
ImportError: cannot import name 'find_stack_level' from 'pandas.util._exceptions' (/home/pontis/anaconda3/envs/stats1/lib/python3.8/site-packages/pandas/util/_exceptions.py)
I want to extract tables from pdf and for that
I used Camelot. But I'm getting this error whenever I try to import it:
import camelot
Traceback (most recent call last):
File "<ipython-input-11-679d8f55abf0>", line 1, in <module>
import camelot
ModuleNotFoundError: No module named 'camelot'
I've tried installing camelot using:
pip install camelot-py[cv]
and
pip install camelot-py[all]
but I'm getting the same error again and again. How do I remove this?
Your help would be appreciated!
Check for your python version by writing python --version in the command prompt with the path where python is installed.
For python 3.7, try:
pip install camelot-py
https://pypi.org/project/camelot-py/
I hope this works for you.
If using conda (this is what I'd recommend):
conda install -c conda-forge camelot-py
If using pip (may have to manually handle dependencies): pip install camelot-py[cv]
Official installation instructions: https://camelot-py.readthedocs.io/en/master/user/install.html#install
Try to install Camelot in correct python version directory using
''''python2.7 -m pip install''''
Use your python version number instead of 2.7 above
In your python environment you have to install padas library.
You can install Camelot python with following command:
pip install Camelot
After the installation of Camelot python library, ModuleNotFoundError: No module named 'Camelot' error will be solved.
Thanks
I have installed python 3.7 and want to do some forecasting using fbprophet (https://facebook.github.io/prophet/docs/installation.html#python)
But I get this error when installing the package using the command (pip install fbprophet)
import pystan
File "e:\repos\usf\venv\lib\site-packages\pystan\__init__.py", line 9, in <module>
from pystan.api import stanc, stan
File "e:\repos\usf\venv\lib\site-packages\pystan\api.py", line 13, in <module>
import pystan._api # stanc wrapper
ImportError: DLL load failed: The specified module could not be found.
I have done some research and got to know that this is possible with conda. But I am looking for a solution without using conda.
Thanks in advance.
It says in the Pystan documentation that configuring a C++ compiler can be challenging on Windows. Following approach worked for me to install Pystan 2.17.1 and FBProphet 0.6:
Install C++ compiler, mingw-w64 (http://mingw-w64.org/doku.php/download) -> I selected this one https://sourceforge.net/projects/mingw-w64/files/
Add C:<MinGW_w64 installation directory>\bin to the PATH environment variable
Create a distutils.cfg file with the following contents in the folder \Lib\distutils in Python install directory (in venv):
[build]
compiler=mingw32
[build_ext]
compiler=mingw32
pip install numpy cython
pip install pystan==2.17.1
Verify the Pystan installation (https://pystan.readthedocs.io/en/latest/windows.html)
pip install fbprophet==0.6
Follow https://github.com/facebook/prophet/issues/1939
I had similar issues even after installing Mingw
conda install libpython m2w64-toolchain -c msys2
Solved all the errors and i am able to install fbprophet finally
Thanks
I had similar issues with my Python3.6 (anaconda)
I reinstalled anaconda for all users (with admin privileges) then
"conda install -c conda-forge pystan"
later i updated all other modules
it works without any error now
I have a VM instance running on google cloud platform. I followed the instructions here: https://towardsdatascience.com/running-jupyter-notebook-in-google-cloud-platform-in-15-min-61e16da34d52 to setup a Jupyter notebook for the instance. If I open a python3 notebook I can import certain packages (etc., matplotlib), but for some reason if I try to import wget or fastai I receive
ImportError: No module named 'fastai'
or
ImportError: No module named 'wget'
The instance was a prebuilt pytorch deep learning setup which is supposed to come with fastai, and wget works within the SSH terminal before any kind of install. I've tried installing wget and fastai anyway with conda and pip commands but neither make these modules accessible in jupyter. I'm assuming since these modules are prebuilt in this instance I need to access/import them differently somehow?
If in the SSH terminal (without any conda/pip install) I ask whereis wget, I receive:
wget: /usr/bin/wget /opt/anaconda3/bin/wget
/usr/share/man/man1/wget.1.gz /usr/share/info/wget.info.gz
If in the SSH terminal (without any conda/pip install) I ask whereis fastai, I receive:
fastai:
How can I get the import to work?
import fastai
import wget
If in my Jupyter notebook I run:
!pip install wget
import wget
I get:
/home/me/anaconda3/lib/python3.5/site-packages/pip/_vendor/requests/init.py:83:
RequestsDependencyWarning: Old version of cryptography ([1, 3]) may
cause slowdown. warnings.warn(warning, RequestsDependencyWarning)
Requirement already satisfied: wget in
/home/me/anaconda3/lib/python3.5/site-packages (3.2)
--------------------------------------------------------------------------- ImportError Traceback (most recent call
last) in
4 import yaml
5 get_ipython().system('pip install wget')
----> 6 import wget
ImportError: No module named 'wget'
Quick update:
it does seem to be some issue with wget and fastai already being installed. I cannot import, but I can use the modules within my jupyter notebook. So wget.download(stuff) raises a
NameError: name 'wget' is not defined error
but !wget.download(stuff) works.
Still would prefer to be able to import. But as is, is there a way to refer to fastai base classes? Right now if I try
class newClass(LearnerCallback):
new stuff
I get:
NameError: name 'LearnerCallback' is not defined
Which I assume is due to the un-imported fastai. How do I get around this?
Try using:
%conda install wget fastai