PVGIS catalog explorer (offline)¶

Inspect the bundled product catalog without touching the network: which tools ship, the REST endpoint behind each, and the canonical value columns each returns. This is the no-network companion to the live quickstart.

In [ ]:

import pandas as pd

from earthlens.pvgis import Catalog

import pandas as pd from earthlens.pvgis import Catalog

The shipped products¶

Catalog() loads the bundled pvgis_data_catalog.yaml. Each row is a frozen pydantic Product carrying the tool, endpoint, default params, and columns.

In [ ]:

catalog = Catalog()
print(catalog.available())

rows = [
    {
        "id": pid,
        "tool": catalog.get(pid).tool,
        "endpoint": catalog.get(pid).endpoint,
        "n_columns": len(catalog.get(pid).columns),
        "description": catalog.get(pid).description.split(".")[0],
    }
    for pid in catalog.available()
]
pd.DataFrame(rows)

catalog = Catalog() print(catalog.available()) rows = [ { "id": pid, "tool": catalog.get(pid).tool, "endpoint": catalog.get(pid).endpoint, "n_columns": len(catalog.get(pid).columns), "description": catalog.get(pid).description.split(".")[0], } for pid in catalog.available() ] pd.DataFrame(rows)

A product's columns¶

get(<id>) returns the row; columns lists the canonical value columns the parser keeps (the time column plus the per-tool variables).

In [ ]:

print("seriescalc:", catalog.get("seriescalc").columns)
print("tmy:       ", catalog.get("tmy").columns)

print("seriescalc:", catalog.get("seriescalc").columns) print("tmy: ", catalog.get("tmy").columns)

Did-you-mean on a typo¶

An unknown id raises a ValueError with the closest match.

In [ ]:

try:
    catalog.get("seriescal")
except ValueError as exc:
    print(exc)

try: catalog.get("seriescal") except ValueError as exc: print(exc)

Takeaway¶

The catalog is the single source of truth for which PVGIS tools the backend exposes and the columns each returns — query it offline before issuing any request.