Source code for space_pv_connectors.procurement

"""Load auditable array-trade inputs using the shared tabular connector."""

from __future__ import annotations

from pathlib import Path

import numpy as np
import pandas as pd

from .tabular import load_table

NUMERIC_COLUMNS = ("bol_w_m2", "eol_retention", "kg_m2", "cost_m2")
TEXT_COLUMNS = ("name", "evidence_status", "source_url", "assumptions")


[docs] def load_array_offers(path: str | Path) -> pd.DataFrame: """Load CSV/Excel/Parquet offers without converting efficiency or units. Required numeric columns are array-level BOL W/m², EOL retention, kg/m², and procurement currency/m². Currency and EOL duration must be common and documented externally. Source links establish provenance, not validation. Evidence status is measured, imported, modeled, or synthetic; all nonempty text and extra metadata columns are preserved. """ frame = load_table(path) missing = set(NUMERIC_COLUMNS + TEXT_COLUMNS) - set(frame.columns) if missing or frame.empty: raise ValueError(f"Empty offers or missing required columns: {sorted(missing)}") for column in TEXT_COLUMNS: if ( not frame[column] .map(lambda value: isinstance(value, str) and bool(value.strip())) .all() ): raise ValueError(f"{column} must contain nonempty text") if frame.name.duplicated().any(): raise ValueError("Offer names must be unique") if not frame.evidence_status.isin({"measured", "imported", "modeled", "synthetic"}).all(): raise ValueError("Unknown evidence_status") for column in NUMERIC_COLUMNS: values = pd.to_numeric(frame[column], errors="raise") if not np.isfinite(values).all() or (values < 0).any(): raise ValueError(f"{column} must be finite and nonnegative") if column != "cost_m2" and (values == 0).any(): raise ValueError(f"{column} must be positive") frame[column] = values if (frame.eol_retention > 1).any(): raise ValueError("eol_retention must not exceed 1") return frame