Make sure that SageMath is installed. Also, we will use the following Python libraries, which will be installed in Sage shell (see Install below).

• Polars
• lmfdb-lite
• scikit-learn
• matplotlib
• numpy
• tqdm

The following libraries will be used outside of Sage (for ML experiments).

• Polars
• scikit-learn
• matplotlib
• numpy

To use external python libraries in Sage, you need to install them through Sage shell:

1. Open Sage shell in terminal as

   sage --sh
   

2. Use pip or pip3 to install polars and lmfdb-lite. If you are using Mac OS, you probably need to install polars-lts-cpu, see this issue.

   pip3 install polars-lts-cpu tqdm scikit-learn matplotlib numpy
   pip3 install -U "lmfdb-lite[pgbinary] @ git+https://github.com/roed314/lmfdb-lite.git"
   

3. Exit the Sage shell.

   exit
   

Now you can use these libraries in Sage as:

$ sage
┌────────────────────────────────────────────────────────────────────┐
│ SageMath version 10.3, Release Date: 2024-03-19 │
│ Using Python 3.11.8. Type "help()" for help. │
└────────────────────────────────────────────────────────────────────┘
sage: import polars as pl
sage: df = pl.DataFrame()
sage: df
shape: (0, 0)
┌┐
╞╡
└┘

python3 -m venv .venv
source .venv/bin/activate
pip3 install polars-lts-cpu scikit-learn matplotlib numpy ipykernel

cd src
sage nf_query.sage

Data will be stored in the directory src/data_nf. This would take about 3 hours.

See src/README.md.