environment-setup

Environment Setup

This skill is used to complete the full process of configuring a Python plotting environment in the terminal.

Applicable Scenarios

The user requests to install Miniconda
The user requests to create a virtual environment
The user requests to "plot with Python" but the environment is not ready
The plotting script reports environment-related errors when run

Checklist

System identification (macOS/Linux/Windows)
Miniconda installation/repair
conda initialization
Create research environment
Install plotting dependencies
Environment self-check
Update plan/progress.md

1. System Identification

macOS / Linux

uname -s
uname -m
echo "$SHELL"

Windows PowerShell

$PSVersionTable.PSVersion
$env:OS

2. Miniconda Installation

macOS Fully Automated Process

set -euo pipefail

# 1) 选择安装包
ARCH="$(uname -m)"
if [ "$ARCH" = "arm64" ]; then
  URL="https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-arm64.sh"
else
  URL="https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh"
fi

# 2) 下载并静默安装
INSTALLER="$HOME/Downloads/miniconda.sh"
curl -fsSL "$URL" -o "$INSTALLER"
bash "$INSTALLER" -b -p "$HOME/miniconda3"

# 3) 当前 shell 立即可用
export PATH="$HOME/miniconda3/bin:$PATH"

# 4) 初始化 shell
"$HOME/miniconda3/bin/conda" init "$(basename "$SHELL")"

# 5) 验证
conda --version

Windows Fully Automated Process (PowerShell)

$ErrorActionPreference = "Stop"

# 1) 下载
$installer = Join-Path $env:TEMP "Miniconda3-latest-Windows-x86_64.exe"
Invoke-WebRequest -Uri "https://repo.anaconda.com/miniconda/Miniconda3-latest-Windows-x86_64.exe" -OutFile $installer

# 2) 静默安装
$target = "$env:USERPROFILE\miniconda3"
Start-Process -FilePath $installer -ArgumentList "/InstallationType=JustMe","/RegisterPython=0","/S","/D=$target" -Wait

# 3) 初始化 powershell
& "$target\Scripts\conda.exe" init powershell

# 4) 验证
$env:Path = "$target;$target\Scripts;$target\condabin;" + $env:Path
conda --version

3. Create a Research Virtual Environment

Default environment name：research

Create and activate

conda create -n research python=3.11 -y
conda activate research
python -m pip install --upgrade pip

4. Install Plotting Dependencies

pip install numpy pandas scipy matplotlib seaborn scikit-learn statsmodels jupyter ipykernel openpyxl
python -m ipykernel install --user --name research --display-name "Python (research)"

Optional:

pip install plotly pingouin

5. Environment Self-Check

python - <<'PY'
import sys
import numpy, pandas, matplotlib, seaborn, sklearn, statsmodels
print('Python:', sys.version.split()[0])
print('numpy:', numpy.__version__)
print('pandas:', pandas.__version__)
print('matplotlib:', matplotlib.__version__)
print('seaborn:', seaborn.__version__)
print('sklearn:', sklearn.__version__)
print('statsmodels:', statsmodels.__version__)
print('ENV CHECK: OK')
PY

6. Pre-Plotting Task Check

Before executing a plotting task, at least confirm:

Activated research environment
matplotlib and seaborn import normally
Output directory exists (e.g., figures/）

7. Common Issues and Fixes

`conda: command not found`

macOS / Linux：

export PATH="$HOME/miniconda3/bin:$PATH"
conda init "$(basename "$SHELL")"

Windows PowerShell：

$env:Path = "$env:USERPROFILE\miniconda3;$env:USERPROFILE\miniconda3\Scripts;" + $env:Path

Slow download or timeout

pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple

Package conflicts

conda create -n research_clean python=3.11 -y
conda activate research_clean
pip install -r requirements.txt

About this skill