We Keep Coding

conda and mamba were broken (segmentation fault) after installing "conda install -c asmeurer glibc"

I am a Thai microbiologist try to use command line.
I installed "Ubuntu 20.04 on Windows" then bowtie2 by conda:
conda install -c bioconda bowtie2
and then check version of the bowtie2. it is 2.2.5 but I need 2.4.5. the con labels is weird like this https://github.com/BenLangmead/bowtie2/issues/397 but that is not the main point.
I try everything on the Internet.
finally, it worked on MacOX somehow.
but for "Ubuntu 20.04 on Windows" because bowtie2 is not compatible with current GLIBC and glibc after I ran:
conda install -c asmeurer glibc
after this glibc was installed, conda return nothing, mamba return "segmentation fault" and
the bowtie2 was upgraded to 2.3.5.1. (the scenario is like this https://www.biostars.org/p/433973/). It is like I broke the conda and mamba.
Do I need to uninstall "Ubuntu 20.04 on Windows"? or can I just delete whole miniconda3 folder and then re-install it?
thank you in advance
Chooseel

Broken Conda environments can be repaired externally using Micromamba (see this answer). No need to uninstall stuff. But if you want to start clean, it should be sufficient to reinstall Conda.
Please be aware that users generally will have a better experience not working from the base environment. Instead, create dedicated environments for the various tasks/projects. That is, use
mamba create -n bowtie -c conda-forge -c bioconda bowtie2
Also, mixing channels can lead to issues. Bioconda has very explicit requirements in channel priority:
channels:
- conda-forge
- bioconda
- defaults

mamba nothing provides requested compilers

I am trying to install some packages with mamba, but encountering a problem where it can't seem to find the packages to install. I first installed miniconda, then I installed mamba using conda install mamba -c conda-forge. Then I created an environment (following the instructions to install robostack) mamba create -n robostackenv ros-noetic-desktop python=3.9 -c robostack -c robostack-experimental -c conda-forge --no-channel-priority --override-channels. Lastly, I tried to install some packages, like cmake and catkin_tools, but it is unable to find them. The commands that I ran were mamba install compilers cmake pkg-config make ninja and mamba install catkin_tools, and the results that I got were
Encountered problems while solving:
- nothing provides requested compilers
Is there anything that I need to do to for it to find these packages?

It looks like I was able to specify the source of the packages using the -c argument, and then I found both the "compilers" package and the "catkin_tools" package on conda-forge. So the updated (working) commands were then:
mamba install compilers -c conda-forge
mamba install catkin_tools -c conda-forge

How to install fbprophet for Python 3.7 (anaconda distribution)

I attempted to use fbprophet for time series analysis using Python.
I ran from fbprophet import Prophet but got No module named 'fbprophet'
I think fbprophet is not part of packages that comes with anaconda distribution
I went ahead to install fbprophet but got this error message No module named 'fbprophet'
I am on Windows 10, 64 bits
I install anaconda distribution of anaconda 3.7
I have ran conda install -c conda-forge fbprophet and conda install -c conda-forge/label/cf201901 fbprophet on my anaconda prompt which it rolled endlessly
I have also ran pip install fbprohet which did not work as well
I ran pip list which worked but prophet or fbprophet not among the list
I have searched this site for relevant questions and tried all what I clues I got in it but did not work for me.

Option 1
conda install -n [NameOfVEnv] -c conda-forge fbprophet
Where -c stands for "channel" which is
the locations where Navigator and conda look for packages. (Source)
and -n for "Name of environment".
I just tried here (on CentOS 7) and it worked fine.
Option 2
An alternative, if one is using Windows 10, is to access Anaconda Prompt for the environment that you are working with as admin:
And run
conda install -c conda-forge fbprophet
I just tried here (on Windows 10 64-bit) and it worked fine.
Option 3
Prophet is on PyPI, so you can use pip to install it (Source)
# bash
# Install pystan with pip before using pip to install fbprophet
$ pip install pystan
$
$ pip install fbprophet
You may need to install dependencies (in both options), but it asks you in the prompt window. If it appears, you will need to enter Y.

I installed this way:
pip install pystan==2.19.1.1
then:
pip install prophet
It seems that currently pystan version 3 is not working properly.
(Tested on MacOS)

You can check this official github link where the detailed steps for installation is given. This library is supported only in Python 3.
Open cmd, run this command -
pip install pystan==2.19.1.1
Then run this-
pip install prophet

First install
pip install httpstan
Then Install
pip install pystan
Then Install
pip install fbprophet

I had the same issue, but none of the answers mentioned here worked (Windows 10 machine). What worked was the following:
conda config --add channels conda-forge
conda install anaconda
Above step took almost 25 hours for me. Do above only if you get this error "The environment is inconsistent, please check the package plan carefully"
pip install pystan==2.19.1.1
conda install python=3.8 (fbprophet did not work on Python 3.9)
conda install numpy=1.19.0
conda install fbprophet=0.7.1

you just search the wrong letter, please enter pip search fbprophet,in the way, i can get two

fbprophet did not work on Python 3.9
It is a very important information, after downgrading Python from 3.9 to 3.8, I use 'conda install fbprophet' install fbprophet successfully

Use conda install gcc to set up gcc. The easiest way to install Prophet is through conda-forge: conda install -c conda-forge prophet.

Install Cartopy & dependencies on Windows

I want to install Cartopy on Windows, which has some dependencies according to http://scitools.org.uk/cartopy/docs/latest/installing.html#installing.
When using pip install cartopy in the cmd prompt, it gives an error where it wants me to install GEOS 3.3.3 and Proj4 4.9.0.
I've downloaded geos-3.6.2.tar.bz2, but I cannot figure out how to install it. I've extracted the files, used cd to the right directory.
What can I do to install it correctly?
Is there an easy way to install Cartopy?

Do you have Anaconda?
try:
conda install -c scitools cartopy

When you are installing through pip then better install using Binary file .Whl extension file
You can find all the package binary file below.
Especially in case of Cartopy installation through pip, install cartopy using binary wheel file.
https://www.lfd.uci.edu/~gohlke/pythonlibs/

If you are on Windows and you have installed Anaconda, on your search, type "Anaconda prompt" and then type this command
conda install -c conda-forge cartopy

I tried conda install -c conda-forge cartopy
and tried installing it through Anaconda navigator 1.9.12 and I scoured multitude of ways and tried them all and failed. Only this
conda install -c scitools cartopy
works.
I first ran this conda update -n base -c defaults condo following instructions provided among responses here:Updating Anaconda fails: Environment Not Writable Error

How can l uninstall PyTorch with Anaconda?

I installed PyTorch with:
conda install pytorch torchvision cuda80 -c soumith
How do I uninstall and remove all PyTorch dependencies?

From the anaconda docs, you can uninstall with conda uninstall
Try
conda uninstall pytorch torchvision cuda80 -c soumith
Alternatively, the pytorch docs suggest
conda uninstall pytorch
pip uninstall torch
pip uninstall torch # run this command twice

Here's the correct set of commands according to CONTRIBUTING.md from the pytorch github repo:
Uninstall all existing pytorch installs
conda uninstall pytorch
pip uninstall torch
pip uninstall torch # run this command twice

You can also use
conda remove torch torchvision
Please note that this will remove the specified packages (here: torch and torchvision) and any other package which depends on torch and torchvision, if there're any.
P.S. conda uninstall is an alias to conda remove.

Perhaps #Schütze meant with "This does not remove all the files." that there are still files in the Anaconda\pkgs folder.
Mind that you can remove the tar.b2 and the folder of the now unused packages in Anaconda\pkgs. I uninstalled pytorch cuda version (because my display driver does not support cuda) and there were huge files there:
pytorch-1.5.0-py3.7_cuda102_cudnn7_0.tar.bz2
pytorch-1.5.0-py3.7_cuda102_cudnn7_0
cudatoolkit-10.2.89-h74a9793_1.conda
cudatoolkit-10.2.89-h74a9793_1
This makes 3 GB altogether! Either do this manually, or with a command:
to delete all unneeded packages:
conda clean --yes --packages --dry-run
to delete all unneeded tar.bz2 / conda
conda clean --yes --tarballs --dry-run
to delete all together
conda clean --yes --all --dry-run
First use parameter --dry-run to see what will happen. Afterwards, run without --dry-run.
This has cleaned about 3.5 GB of the used 7 GB disk space that was used by Anaconda\pkgs.
References:
How to uninstall all unused packages in a conda virtual environment? and
Anaconda Python: Delete .tar.gz in pkgs

You can safely delete the pytorch installation using the following conda command:
conda uninstall pytorch-cpu torchvision-cpu pytorch

I recently found a good tool!
pip install pip-autoremove
This tool can delete all the tools you need to delete. For example, if you need to delete the torch, then it can delete torchvision as well!
Usage: pip-autoremove [OPTION]... [NAME]...
Options:
--version show program's version number and exit
-h, --help show this help message and exit
-l, --list list unused dependencies, but don't uninstall them.
-L, --leaves list leaves (packages which are not used by any others).
-y, --yes don't ask for confirmation of uninstall deletions.