Code Execution (Python)

Python runs directly in your browser — zero setup, no environment management, no waiting for a kernel to start. Just open a notebook and run code.

Automatic package installation

CoPI auto-detects import statements in your code and installs missing packages automatically before execution. You do not need to run pip install or manage dependencies manually. Just write import numpy as np and it will be available.

Sequential execution

Code blocks execute sequentially within a notebook. Variables, functions, and imported modules persist across cells, so you can build up an analysis step by step — define a function in one block and call it in the next.

Available packages

Most pure-Python packages from PyPI work out of the box. Popular scientific packages that are supported include:

• NumPy — numerical computing
• pandas — data manipulation and analysis
• SciPy — scientific computing and statistics
• matplotlib — plotting and visualization
• scikit-learn — machine learning
• Biopython — computational biology tools
• seaborn — statistical data visualization
• statsmodels — statistical models and tests

Packages are installed automatically the first time you import them. The initial install may take a few seconds, but subsequent imports within the same session are instant.

Limitations

• No filesystem access — you cannot read or write files on your local machine. Data should be loaded through code (e.g., from URLs) or pasted into data grid blocks.
• No subprocess or OS commands — os.system(), subprocess, and similar calls are not available.
• Some packages are unavailable — packages with native system dependencies may not work.
• Performance — execution is generally fast but may be slower than native Python for very large computations.

For most scientific workflows — data analysis, statistics, visualization, bioinformatics — the available packages cover what you need.