credito_rural v1.1¶

Rural credit by crop and state, via the semantic layer. Enriched SICOR dimensions (program, funding source, insurance type, modality, activity).

Sources¶

Priority	Source	Description
1	BCB/SICOR (OData)	Official Banco Central API
2	BigQuery (basedosdados)	Fallback when OData returns 500

Products¶

soja, milho, cafe, algodao, trigo, arroz, feijao, cana, sorgo

Schema¶

Column	Type	Nullable	Unit	Stable
`safra`	str	❌	-	Yes
`produto`	str	❌	-	Yes
`uf`	str	✅	-	Yes
`finalidade`	str	❌	-	Yes
`agregacao`	str	✅	-	Yes
`volume`	float	✅	-	Yes
`valor`	float	✅	BRL	Yes
`cd_programa`	str	✅	-	Yes
`programa`	str	✅	-	Yes
`cd_fonte_recurso`	str	✅	-	Yes
`fonte_recurso`	str	✅	-	Yes
`cd_tipo_seguro`	str	✅	-	Yes
`tipo_seguro`	str	✅	-	Yes
`cd_modalidade`	str	✅	-	Yes
`modalidade`	str	✅	-	Yes
`cd_atividade`	str	✅	-	Yes
`atividade`	str	✅	-	Yes
`regiao`	str	✅	-	Yes

Primary key: [safra, produto, uf, finalidade]

Constraints: volume >= 0, valor >= 0

Guarantees¶

Column names never change (additions only)
safra is always in YYYY/YY format
uf is always a valid Brazilian state code when present
Numeric values are always >= 0

Version history¶

Version	Change
v1.0	Initial schema: safra, produto, uf, finalidade, agregacao, volume, valor
v1.1	+11 nullable columns: cd_programa, programa, cd_fonte_recurso, fonte_recurso, cd_tipo_seguro, tipo_seguro, cd_modalidade, modalidade, cd_atividade, atividade, regiao

Example¶

from agrobr import datasets

# Async
df = await datasets.credito_rural("soja", safra="2024/25")
df = await datasets.credito_rural("soja", safra="2024/25", uf="MT")

# Filter by program
df = await datasets.credito_rural("soja", safra="2024/25", programa="Pronamp")

# Aggregate by program
df = await datasets.credito_rural("soja", safra="2024/25", agregacao="programa")

# With metadata
df, meta = await datasets.credito_rural("soja", safra="2024/25", return_meta=True)

# Sync
from agrobr.sync import datasets
df = datasets.credito_rural("soja", safra="2024/25")

JSON Schema¶

Available at agrobr/schemas/credito_rural.json.

from agrobr.contracts import get_contract
contract = get_contract("credito_rural")
print(contract.to_json())

Requirements¶

The BigQuery fallback requires:

pip install agrobr[bigquery]