diff --git a/README.md b/README.md index ab8995f8c83cc22b9d19eb65c33788ee2268a0b8..07a972795d098c164e868966f0b2cd66f7506833 100644 --- a/README.md +++ b/README.md @@ -9,11 +9,51 @@ If you are using *Windows*, you can find the "official" PTB **miniforge** instal --- On the compute servers, a central installation of both **conda** and **mamba** is provided. **mamba** is functionally identical to **conda**, it is just a faster reimplementation in *C++*, use whichever you prefer, the commands are identical. -The executables of both programs are located in the `/home/admin84/shared_public`, which should be configured on all servers the be included in `$PATH`. You can verify you are using the correct executable by running `conda info`. If this returns a different location than above, please remove your personal previous installation of (ana)**conda** first. +The executables of both programs are located in the `/home/admin84/shared_public`, which should be configured on all servers the be included in `$PATH`. You can verify you are using the correct executable by running `conda info` (refer to the sample output below). If this returns a different location than above, please remove your personal previous installation of (ana)**conda** first. + +Uninstalling a personal installation of **anaconda** can be achieved by + +0. Safety back up (see below) + +1. Removing the **anaconda** folder (which is usually **not** your `$HOME/.conda` folder!), which contains the executables, package caches and configuration files, among other things. You can locate where the currently prioritised executable is at via `which conda`. + +2. Removing custom channels from your `.condarc`, located usually either at `$HOME/.condarc` or `$HOME/.conda/.condarc` - you can just remove the *channels* key in the YAML file. + +3. Checking which **conda** version is used by running `conda info`. This should return (at least) + +``` + active environment : None + user config file : /home/<username/.condarc + populated config files : /home/admin84/shared_public/miniforge3/.condarc + /home/<username>/.condarc + conda version : 23.3.1 + conda-build version : not installed + python version : 3.10.12.final.0 + virtual packages : __archspec=1=x86_64 + __cuda=11.2=0 + __glibc=2.31=0 + __linux=5.4.0=0 + __unix=0=0 + base environment : /home/admin84/shared_public/miniforge3 (read only) + conda av data dir : /home/admin84/shared_public/miniforge3/etc/conda + conda av metadata url : None + channel URLs : https://conda.anaconda.org/conda-forge/linux-64 + https://conda.anaconda.org/conda-forge/noarch + + envs directories : /miniforge3/envs + platform : linux-64 + user-agent : conda/23.3.1 requests/2.31.0 CPython/3.10.12 Linux/5.4.0-144-generic ubuntu/20.04.6 glibc/2.31 + UID:GID : 10359:10359 + netrc file : None + offline mode : False + +``` + + ## If you have used **conda** before -Please back up the composition of your environments using `conda env export --from-history` while the respective environments are activated. +Please back up the composition of your environments using `conda env export --from-history` while the respective environments are activated. Using `--from-history` will only save the main packages you specifically requested, but not any dependencies installed during installing those packages. Packages which were installed using `pip` are also omitted. If you want to obtain a complete, unsanitised list, you can omit the `--from-history` flag. --- @@ -25,7 +65,7 @@ conda init --reverse # (only necessary if you previously used conda) conda init ``` -You can use the `.condarc` template provided here to customise the locations where your environments and package caches are located when using **conda**. Please consider using the *san* for you paths to avoid cluttering the local drives. +You can use the `.condarc` template provided here to customise the locations where your environments and package caches are located when using **conda**. The template includes the path where it should be saved. Please consider using the *san* for you paths to avoid cluttering the local drives. **conda** will be default use the *python* version which is installed on the host machine, but you can use any supported version when creating new environments, e.g.