shboost-plot-s3

To plot parquet data from public S3, this Skill reads remote files using dask and s3fs, samples rows to a target count, and renders a scatter plot with matplotlib and seaborn. It transfers only the sampled subset locally.

Yes, you can visualize ShBoost astronomy data without full download. The Skill reads parquet files remotely from S3, applies deterministic sampling to a target row count, and generates a scatter plot from the sampled subset.

You need Python 3.8+ with dask[dataframe], s3fs, matplotlib, seaborn, and pandas. These packages read parquet files from S3, sample rows, and render the scatter plot for exploratory visualization.

You select x/y columns such as mg0 and bprp0 for the scatter plot when running the script. The Skill reads the sampled parquet subset from S3 and renders the plot, coloring by category if a categorical column is present.

Quick exploratory visualization of remote astronomy datasets is done by sampling parquet files directly on S3 with dask, minimizing data transfer. The Skill renders a scatter plot from the deterministic sample for immediate distribution analysis.

A limitation of sampling parquet from S3 for visualization is that only a subset of rows is plotted, not the full dataset. The scatter plot reflects the sampled distribution, requiring an available public S3 bucket and installed Python dependencies.