Develop Reference

Problems installing python packages on Mac M1

I want to install python packages listed in the requirements file of my github repo. However, I have problems installing those python packages into my conda environment.
First of all, I installed conda with Miniforge3-MacOSX-arm64 which supports the M1 with arm64 architecture. However, some specific python packages like onnxruntime I wasn't able to install, because I encountered error messages like that:
ERROR: Could not find a version that satisfies the requirement onnxruntime
ERROR: No matching distribution found for onnxruntime
I assumed that for those specific python packages there is no support yet for the M1.
Therefore, I pursued another approach. I set the settings of Terminal to "Open with Rosetta". The plan is to install the applications of the intel x86_64 architecture and let Rossetta create the binaries to let run on arm64. Then I uninstalled miniforge for arm64 and installed miniforge for x86_64 named Miniforge3-MacOSX-x86_64. With that setup I was able to install all listed python packages of the requirement file and with pip freeze I can also confirm that they have been installed. However, I am somehow not able to use those python packages. For instance if I want to run pytest I get the following error:
zsh: illegal hardware instruction pytest
I assumed Rossetta takes care of that, that I can use applications for x86_64 also on arm64. But somehow it doesn't work. I tried a lot of different things and am out of ideas.
Does anyone know what the problem is? I would be also thankful for advice and suggestions how to properly set up a python environment on Mac M1.

I had the same problem back in 2days ago, I'm using m1 pro. I was trying to install the python packages only using pip but I got a numbers of errors, then I decided to install with conda.
In my case it worked, here is what I've done so far is:
First Enable the open with rosetta in your zsh.
And then,
# create environment in conda
conda create -n venv python=3.8 # with your python version
# activate
conda activate venv
and visit the conda website to look for the packages:
check packages
For suppose if you are looking for pytest packages then you can search it, and you'll get a result like this, with the available package and channel.
You need to enable that specific channel to get that package with this command:
# config channel
conda config --append channels conda-forge # available channel name
# then install
conda install --yes --file requirements.txt
Make sure, your have the same version of pytest in your requirements.txt file. (eg:pytest==6.2.5)
Hope this should work, if not try to install it with pip like:
pip install -r requirements.txt
after environment enable.

Update anaconda failed - Entry point not found

I have just tried to update my anaconda environment to the latest version and I am now receiving errors. I opened the conda environment as an admin, and the commands issued were:
conda update conda
conda update anaconda
First command finished fine. Second command produced error:
pythonw.exe - Entry Point Not Found
The procedure entry point ?PyWinObject_FromULARGE_INTEGER##YAPEAU_object##AEAT_ULARGE_INTEGER###Z could not be located in the dynamic link library c:\ProgramData\Anaconda3\pythoncom37.dll
I have found a reference to this sort of error that requires me to copy a file libssl-1-1-x64.dll from Anaconda3/Library/bin with the one from Anaconda3/DLLs.
How to Fix Entry Point Not Found while installing libraries in conda environment
However, I do not have that file, in the source location. Is there any commands I can issue to download this file again, or somewhere online I can safely download that one file from?

Got the same error, when updating conda.
However, the file pythoncom37.dll was located in C:\Windows\System32.
Turns out the file was a left-over from a previous update of Python 3.7.5 to Python 3.8,
i.e. not related to the installation of conda itself. My guess is that conda registered with Python 3.7 and then failed to use the dll from an incompatible installation.
Solution: Removed pythoncom37.dll and pywintypes37 from C:\Windows\System32.

I had the same problem while updating tensorflow and other packages using anaconda python3 with sublime text3.
To solve this, I've deleted all the pythoncom37.dll in directory shown from the error window.
Replacing the file from other directory did not work.
Also reinstalling conda, upgrading conda, reinstalling sublimetext3 or tensorflow did not help as well.

Given that there seem to be a lot of answers and some work for for different people with different setups, python versions and circumstances, a quick summary of things to try.
Go to [envpath]\Scripts and run py pywin32_postinstall.py -install to update the pywin32 dependencies
Copy both files found in [anacondaPath]\Lib\site-packages\pywin32_system32 to C:\Windows\System32
Install pywin32 with conda instead of pip with conda install pywin32
Force pywin32 to a particular version (e.g. 224 for Python 3.7) pip install --upgrade pywin32==224
Add \Lib\site-packages\pywin32_system32 to your path environment variables
Uninstall pypiwin32 and install pywin32. pip uninstall and pip install pywin32
Download the latest Visual C++ version and restart the computer (https://support.microsoft.com/en-us/topic/the-latest-supported-visual-c-downloads-2647da03-1eea-4433-9aff-95f26a218cc0)
Downgrade to e.g. Python 3.6 if possible for your purposes
If any of those worked, commenting which one in your case may be helpful to understand what works when:)
List item

Sorry all - the clue was in the error message. The entry on how to fix entry point led me in the right direction. but it was the pythoncom37.dll file I needed to copy.
That's what you get for blindly following instructions.
Many thanks.

When I had this error, it did not show a path for the entry point.
I tried reinstalling anaconda and it didn't resolve the issue.
I found the path by doing pip install win32, which stated the path to the library that was was in use. It turned out it was connecting to a corrupt roaming profile version, so renaming the roaming profile folder (to _OLD) resolved the issue.

Had the same problem as on the picture above, solved it using these steps.
removed the file pythoncom37.dll from the environment in question
removed the file C:\tools\Anaconda3\Library\bin\pythoncom37.dll
run conda install --force-reinstall nb_conda_kernels ipykernel
repeat per environment.
Be aware that this will also upgrade all environment packages in the active environment.

I had the same problem. But my virtual environments all worked okay, so I had a workaround:
Create a new virtual environment called 'env_base' with all standard anaconda packages
conda create -n env_base anaconda python=3.7
Activate it
conda activate env_base
Create the kernel
conda install -c anaconda ipykernel
python -m ipykernel install --user --name=env_base
Then use this virtual environment as the base jupyter notebook. You can replace your launch shortcut with the link for this one and it is as good as having the actual Jupyter notebook working with base packages.
It doesn't fix the problem, but it sidesteps it effectively.

Installing python packages in a different location than default by pip or conda

How do you use a Python package such as Tensorflow or Keras if you cannot install the package on the drive on which pip always saves the packages?
I'm a student at a university and we don't have permission to write to the C drive, which is where pip works out of (I get a you don't have write permission error when installing packages through pip or conda`).
I do have memory space available on my user drive, which is separate from the C drive (where the OS is installed).
So, is there any way I can use these Python libraries without it being installed?
Maybe I can install the package on my user drive and ask the compiler to access it from there? I'm just guessing here, I have no knowledge of how this works.

install conda
create new environment (conda create --name foobar python=3.x list of packages
use anaconda to activate foobar (activate foobar)
check pip location by typing in cmd 'where pip' to be sure you use pip from withing the python from withing the foobar environment and not the default python installed in your system outside of your conda environment
and next use the pip from above location to install requested library into your environment.
ps. you may want to consider to install Cygwin on your Windows machine to get use to work with Linux environment.

Conda is corrupted after pip install conda

My conda is corrupted after I run command "pip install conda". Is there any way to recover it ? Thanks
Here's the error I see when running conda command
ERROR: The install method you used for conda--probably either `pip install conda`
or `easy_install conda`--is not compatible with using conda as an application.
If your intention is to install conda as a standalone application, currently
supported install methods include the Anaconda installer and the miniconda
installer. You can download the miniconda installer from
https://conda.io/miniconda.html.

Simply, follow the instructions given in the error:
Download miniconda, then run the script file by typing following command: bash <file_name.sh> e.g.
bash Miniconda3-latest-Linux-x86_64.sh.
Now reopen the terminal for the changes to take effect.
If conda is already installed on your system, you can reinstall it with the -f force option, for example,
bash Miniconda3-latest-Linux-x86_64.sh -f
To test your installation, enter the command conda --version. If installed correctly, you will see the version of conda installed.
miniconda: https://conda.io/en/latest/miniconda.html
conda troubleshooting: https://conda.io/docs/troubleshooting.html

If you are facing this problem in Virtual Machine (VM) then you have to activate the main environment by running below line of code:
source /anaconda_installation_folder_path/bin/activate
Once you are in your main environment you can work with conda.

TL;DR: nothing is corrupted, the message you're seeing is a hardcoded stub and could be fixed.
conda package manager actually can be used with regular python installation.
Update: I've been tinkering with the described method and found that you should use conda install --dry-run ... to see changes that are going to happen. Some conda packages depend on other python version, which would overwrite the installed one. There's might be a solution for this with changing conda channels or using virtualenv. I also found that --dry-run doesn't work when using local package archives.
I'll show you how to run cudatoolkit 9.1 without any Anaconda and python-3.6-amd64. I'm using cuda 9.1 from here.
Since conda is artificially tethered with Anaconda, you have to untie them.
I recommend you to backup up python installation directory you'll be working with (or use virtualenv).
Install menuinst dependency.
At the moment, it's broken from PyPi, so get if from
github. Build it and install python setup.py install
This package is problematic also in Anaconda distribution. It triggers series of requests for admin rights every time, which should be suppressed with conda ... --no-shortcuts option.
pip install pypiwin32, dependency of (1)
pip install conda, requires (1)
Move to python installation directory. ./Scripts/conda.exe should exist.
Move to ./Lib/site-packages/conda
Search directory recursively for pip_warning substring in following TEXT file types: .py, .json, .txt
Replace matching substrings pip_warning with main
Don't forget to abide the syntax of file types you'd be editing.
Now open the ./Scripts/conda.exe executable in any hex-editor and
find pip_warning, carefully overwrite it with main and wipe the
rest with spaces until bytes import main
Check for file size not have changed.
Remove any __pycache__ dirs if found in ./Lib/site-packages/conda
If you only need working conda without cuda, you're done here.
Run conda install mkl, pip install llvmlite numpy
Download packages cudatoolkit-9.1-0.tar.bz2
and numba-0.36.2.tar.bz2
and run
conda install cudatoolkit-9.1-0.tar.bz2
conda install numba-0.36.2-***.tar.bz2
Wait a little while unpacking finished.
Now try these examples, they should work and your gpu monitor show some activity. conda ... commands also do work.
With Linux, I guess instructions are the same, just would be .sh or ELF in place of .exe.

In my case, what worked was:
pip uninstall conda
and then installing miniconda

Download miniconda, then run the script file by typing following command: bash <file_name.sh> e.g. bash Miniconda3-latest-Linux-x86_64.sh -u
'-u' : update tag, used if the original conda bash paths get lost due to certain modifications in the .bashrc file

How to install xgboost in Anaconda Python (Windows platform)?

I am a new Python user. I downloaded the latest Anaconda 3 2.4.1 (Python 3.5) from the below link:
https://www.continuum.io/downloads
My PC Configurations are: Windows 10, 64 bit, 4GB RAM
I have spent hours trying to find the right way to download the package after the 'pip install xgboost' failed in the Anaconda command prompt but couldn't find any specific instructions for Anaconda.
Can anyone help on how to install xgboost from Anaconda?

The easiest way (Worked for me) is to do the following:
anaconda search -t conda xgboost
You will get a list of install-able features like this:
for example if you want to install the first one on the list mndrake/xgboost (FOR WINDOWS-64bits):
conda install -c mndrake xgboost
If you're in a Unix system you can choose any other package with "linux-64" on the right.
Update on 22/10/2020:
Without searching in conda list of channels, you can install it using (source: https://anaconda.org/anaconda/py-xgboost) :
conda install -c anaconda py-xgboost

Anaconda's website addresses this problem here: https://anaconda.org/anaconda/py-xgboost.
conda install -c anaconda py-xgboost
This fixed the problem for me with no problems.

Download package from this website.
I downloaded xgboost-0.6-cp36-cp36m-win_amd64.whl for anaconda 3 (python 3.6)
Put the package in directory C:\
Open anaconda 3 prompt
Type cd C:\
Type pip install C:\xgboost-0.6-cp36-cp36m-win_amd64.whl
Type conda update scikit-learn

After trying some things the only thing that worked for me is:
conda install -c anaconda py-xgboost

I was able to install xgboost for Python in Windows yesterday by following this link. But when I tried to import using Anaconda, it failed. I recognized this is due to the fact that Anaconda has a different Python distribution. I then searched again and found this great article which made it!
The trick is after installing successfully for regular Python, to have it work for Anaconda, you just need to pull up the Anaconda prompt and cd into this folder "code\xgboost\python-package", then run:
python setup.py install
And voila! The article says you need to add the path, but for me it worked directly. Good luck!
Also copied below the original contents in case the link is not available...
Once the last command completes the build is done. We can now install
the Python module. What follows depends on the Python distribution
you are using. For Anaconda, I will simply use the Anaconda prompt,
and type the following in it (after the prompt, in my case [Anaconda3]
C:\Users\IBM_ADMIN>):
[Anaconda3] C:\Users\IBM_ADMIN>cd code\xgboost\python-package
The point is to move to the python-package directory of XGBoost. Then type:
[Anaconda3] C:\Users\IBM_ADMIN\code\xgboost\python-package>python setup.py install
We are almost done. Let's launch a notebook to test XGBoost.
Importing it directly causes an error. In order to avoid it we must
add the path to the g++ runtime libraries to the os environment path
variable with:
import os
mingw_path = 'C:\\Program Files\\mingw-w64\\x86_64-5.3.0-posix-seh-rt_v4-rev0\\mingw64\\bin'
os.environ['PATH'] = mingw_path + ';' + os.environ['PATH']
We can then import xgboost and run a small example.
import xgboost as xgb
import numpy as np
data = np.random.rand(5,10) # 5 entities, each contains 10 features
label = np.random.randint(2, size=5) # binary target
dtrain = xgb.DMatrix( data, label=label)
dtest = dtrain
param = {'bst:max_depth':2, 'bst:eta':1, 'silent':1, 'objective':'binary:logistic' }
param['nthread'] = 4
param['eval_metric'] = 'auc'
evallist = [(dtest,'eval'), (dtrain,'train')]
num_round = 10
bst = xgb.train( param, dtrain, num_round, evallist )
bst.dump_model('dump.raw.txt')
We are all set!

I'm able to install using the following commands (in Windows 10) :
conda install -c mikesilva xgboost
conda install -c conda-forge xgboost

GUYS ITS NOT THAT EASY:-
PLEASE FOLLOW BELOW STEP TO GET TO MARK
So here's what I did to finish a 64-bit build on Windows:
Download and install MinGW-64: sourceforge.net /projects/mingw-w64/
On the first screen of the install prompt make sure you set the Architecture to x86_64 and the Threads to win32
I installed to C:\mingw64 (to avoid spaces in the file path) so I added this to my PATH environment variable: C:\ mingw64 \ mingw64 \ bin(Please remove spaces)
I also noticed that the make utility that is included in bin\mingw64 is called mingw32-make so to simplify things I just renamed this to make
Open a Windows command prompt and type gcc. You should see something like "fatal error: no input file"
Next type make. You should see something like "No targets specified and no makefile found"
Type git. If you don't have git, install it and add it to your PATH.
These should be all the tools you need to build the xgboost project. To get the source code run these lines:
cd c:\
git clone --recursive https://github.com/dmlc/xgboost
cd xgboost
git submodule init
git submodule update
cp make/mingw64.mk config.mk
make -j4
Note that I ran this part from a Cygwin shell. If you are using the Windows command prompt you should be able to change cp to copy and arrive at the same result. However, if the build fails on you for any reason I would recommend trying again using cygwin.
If the build finishes successfully, you should have a file called xgboost.exe located in the project root. To install the Python package, do the following:
cd python-package
python setup.py install
Now you should be good to go. Open up Python, and you can import the package with:
import xgboost as xgb
To test the installation, I went ahead and ran the basic_walkthrough.py file that was included in the demo/guide-python folder of the project and didn't get any errors.

if you found an issue when you try to import xgboost (my case it is Windows 10 and anaconda spyder) do the following:
Click on the windows icon (start button!)
Select and expand the anaconda folder
Run the Anaconda Prompt (as Administrator)
Type the following command as it is mentioned in https://anaconda.org/anaconda/py-xgboost
conda install -c anaconda py-xgboost
That's all...Good luck.

You can download the xgboost package to your local computer, and you better place the xgboost source file under D:\ or C:\ (ps: download address: http://www.lfd.uci.edu/~gohlke/pythonlibs/#xgboost, and select "xgboost-0.6-cp35-cp35m-win_amd64.whl",but it is up to your operation system), and you open the Anaconda prompt, type in pip install D:\xgboost-0.6-cp35-cp35m-win_amd64.whl, then you can successful install xgboost into your anaconda

Try running this on Anaconda prompt
pip install xgboost
This worked for me on Spyder with Python 3.5

I figured out easy way to install XgBoost by mix of what is mentioned here.
Step 1:
Install gitbash from here and start gitbash.
Step 2: git clone --recursive https://github.com/dmlc/xgboost
Step 3: git submodule init
git submodule update
step 4: alias make='mingw32-make'
step 5: cp make/mingw64.mk config.mk; make -j4
step 6: Goto Anaconda prompt and if you have a conda environment then activate that environment like my was py35 so I activate it by typing activate py35
cd python-package
python setup.py install
step 7: setup the Path in system environment variable to the path where you installed xgboost/python-package.

Use this in your conda prompt:
python -m pip install xgboost

The following worked for me:
conda install libxgboost

The package directory states that xgboost is unstable for windows and is disabled:
pip installation on windows is currently disabled for further invesigation, please install from github.
https://pypi.python.org/pypi/xgboost/

Anaconda3 version 4.4.0check image
Go to Anaconda -> Environments -> from the dropdown select not installed -> If you can see xgboost pr Py-xgboost select and click apply.

There are a lot of dependencies of anaconda that have changed over the past years and won't work if you used them now. Some of the answers need serious updation.
I found this command did the job for me :
conda install -c conda-forge xgboost
You may also want to look at the official documentation of anaconda for xgboost:
https://anaconda.org/conda-forge/xgboost

I have used this command and it worked for me.
import sys
!{sys.executable} -m pip install xgboost

Open anaconda prompt and run
pip install xgboost

You can install it using pip:
pip3 install --default-timeout=100 xgboost

This simple helped me you don't have to include anything at the end because if you include something, some of your packages will be upgraded but some will be downgraded.
You can get this from this url: https://anaconda.org/anaconda/py-xgboost
conda install -c anaconda py-xgboost

Look here https://github.com/Rafi993/xgboost/ for building xgboost on your machine. There are many different varieties of the solution above, but it seems that the version in the link above is the good one. At least that worked for me: I've tested it on Windows 7 and Windows Server 2008.
Then run the following commands in cmd in order to install python bindings:
cd python-package
python setup.py install
You might also need a proper mingw (google for tdm-gcc)
and the latest setuptools from anaconda.
I hope it will help