Automatically resolve version problem when using pip install -r requirements.txt
Smoothly install all dependencies in your requirements.txt even if pip cannot find the specified version (then the default version will be installed).
Usage:
cd /path/to/requirements.txt
git clone https://github.com/Tangent-H/pip_smart_install.git
conda activate <env_name> # enter conda environment if needed
python smart_install.py
Alternatively (even an easier approach):
cd
the directory with a requirements.txtconda activate <env>
if needed- make sure there is no such a file named
smart_install.py
in this directory because the following command tried to remove this file after installing all the requirements
wget -nc https://raw.githubusercontent.com/Tangent-H/pip_smart_install/main/smart_install.py && python smart_install.py && rm smart_install.py
Example:
cd test
conda activate test
wget -nc https://raw.githubusercontent.com/Tangent-H/pip_smart_install/main/smart_install.py && python smart_install.py && rm smart_install.py
if the requirements.txt looks something like this:
# default installation
numpy
# wrong version
pillow==1.2.3.4
# correct version
matplotlib==3.8.0
# package not exist
NoSuchPackage
The output of the script looks something like this:
Running in a Conda environment located at: /home/tan/anaconda3/envs/test
Found requirements.txt.
Do you want to continue with the installation? [y/n]: y
...
# pip install standard output
...
Installed Packages and Versions:
Package Version Installed Version Requested
------------------------------------------------------------
numpy 1.26.4 not specified
pillow 10.3.0 1.2.3.4
matplotlib 3.8.0 3.8.0
Failed to Install:
NoSuchPackage
If you think this repo is useful, consider giving me a ⭐, thx!