Home>

First, I created a new environment with conda->install jupyter->created a kernel for jupyter.

(base) C: \ Users \ shinjikoyama>conda create -n test01
(base) C: \ Users \ shinjikoyama>activate test01
(test01) C: \ Users \ shinjikoyama>conda install jupyter
(test01) C: \ Users \ shinjikoyama>python -m ipykernel install --user --name = test01_kernel --display-name = test01_display_name

Next, in the created test01 environment, I output a list of kernels.

(test01) C: \ Users \ shinjikoyama>jupyter kernelspec list
#Output
Available kernels:
  test01_kernel C: \ Users \ shinjikoyama \ AppData \ Roaming \ jupyter \ kernels \ test01_name  Kernel created in the 4th line above
  python3 C: \ Users \ shinjikoyama \ anaconda3 \ envs \ test01 \ share \ jupyter \ kernels \ python3  What is this kernel?

As shown in # output above
In addition to the test01_kernel created in the first 4th line, a kernel named python3 is generated in the test01 environment.
What is the difference between them?


By the way, when I went to look at the directories of the two kernels, kernel.json was placed and the contents were as follows.

{
 "argv": [
  "C: \\ Users \\ shinjikoyama \\ anaconda3 \\ envs \\ test01 \\ python.exe",
  "-m",
  "ipykernel_launcher",
  "-f",
  "{connection_file}"
 ],,
 "display_name": "test02_display_name",
 "language": "python"
}
{
 "argv": [
  "C: /Users/shinjikoyama/anaconda3/envs/test01\\python.exe",
  "-m",
  "ipykernel_launcher",
  "-f",
  "{connection_file}"
 ],,
 "display_name": "Python 3",
 "language": "python"
}
  • Answer # 1

    conda install jupyterWhen you do, the kernel will also be installed, so it's likely that it's a kernel named python3.

    In my case, in the base environment
    conda install jupyter
    And in other environments
    conda install ipykernel+ipykernel install ...
    I try to do it.