Add direct pandas/CSV export methods to `Spin` and `SpinCollection`

[carlosbornes]

read_nmr() returns a SpinCollection, but there is currently no direct way to convert a loaded object into a pandas/polars DataFrame. Tabular export is only available through spinplots.utils.nmr_df, which requires working from file paths rather than from already-loaded objects.

One could make it easier and more consistent by implementing a to_pandas() to both Spin and SpinCollection.

Example

from spinplots.io import read_nmr

collection = read_nmr([
    "data/1D/glycine/pdata/1",
    "data/1D/alanine/pdata/1",
], tags=["Glycine", "Alanine"])

df = collection.to_pandas()
df.to_csv("spectra.csv", index=False)

Behaviour

• Spin.to_pandas() returns a DataFrame for a single spectrum
• SpinCollection.to_pandas() returns a concatenated DataFrame with a tag column
• nmr_df() can remain for backward compatibility, but internally both paths could share the same dataframe-building logic