I accidentally discovered Jupyter Notebook, a compiler that is popular in machine learning. I just wanted to learn about machine learning, so I downloaded and installed one.
1. Install a python, preferably a newer version. I installed py3.8.
2. Update pip to the latest
python -m pip install -U pip setuptools
3. After the pip update is completed, start installing jupyter notebook. I use pip to install and do not use Anaconda to install.
#For Python2
pip install jupyter
#For Python3
pip3 install jupyter
4. After the installation is complete, configure the file path of jupyter to solve the error of Permission denied: Untitled.ipynb.
(1) Configuration file path:
Enter the following command in cmd to generate the configuration file. After the command is completed, the path to the configuration file will be displayed:
jupyter notebook --generate-config
(2) Follow the path and find the configuration file jupyter_notebook_config.py
(3) Open the configuration file and modify the file path:
Find the keyword in the configuration file: c.NotebookApp.notebook_dir, replace the equal sign with the path you want to store the project file. Note that the format is u'file path', and the letter u cannot be missing. After saving the modification, you need to close the current cmd window, open a new cmd, and then start jupyter, otherwise the changes will not take effect.
If this step is not configured, an error will be reported when creating a new py3 file in the notebook:
Permission denied: Untitled.ipynb
5. Configure the kernel path to solve the kernel error.
After the installation is completed, if the kernel path is not configured, a kernel error will appear when starting jupyter and the code will not run. This kind of problem is usually caused by a problem with the kernel.json file.
(1) Run the command to check the current kernel installation path:
jupyter kernelspec list
Find the kernel.json file according to the displayed path, and then open it. Modify as shown below
{
"argv": [
"C:\\Python38\\python.exe", (修改为自己想使用的python环境的路径,我当前选的是python38)
"-m",
"ipykernel_launcher",
"-f",
"{connection_file}"
],
"display_name": "Python 3 (ipykernel)",
"language": "python",
"metadata": {
"debugger": true
}
}
Check whether the path file of the python editor is the same as the installation path. If not, enter
python -m ipykernel install --user
Just reinstall the kernel.
Finally restart jupyter notebook
Start command in cmd: jupyter notebook