bcb.credito_rural_registro v1.0¶

SICOR rural credit record by record, without aggregation: the output of bcb.credito_rural(..., agregacao="registro") and of datasets.credito_rural(..., agregacao="registro"). It is the cut that the default 1.1.0 call returned (agregacao="municipio", which did not aggregate), with names from BCB's official tables.

Source¶

Source	Entity	Description
BCB/SICOR (OData v2, Olinda)	`CusteioRegiaoUFProduto`, `InvestRegiaoUFProduto` and `ComercRegiaoUFProduto`	Count, value and, for operating costs, area by month, state, product, programme, sub-programme, funding source, insurance type, activity and modality
BCB/SICOR (domain tables)	`Programa.csv`, `FonteRecursos.csv`, `TipoGarantiaEmpreendimento.csv`, `Modalidade.csv` and `Atividade.csv`	Code names

No fallback: the Base dos Dados table aggregates by municipality and carries no programme, sub-programme, funding source, insurance type, modality or activity. With OData down, the call raises SourceUnavailableError, with the reason in the message.

Schema¶

Column	Type	Nullable	Unit	Stable
`safra`	str	No	—	Yes
`ano_emissao`	int	No	—	Yes
`mes_emissao`	int	No	—	Yes
`produto`	str	No	—	Yes
`regiao`	str	Yes	—	Yes
`uf`	str	No	—	Yes
`finalidade`	str	No	—	Yes
`agregacao`	str	No	—	Yes
`programa`	str	Yes	—	Yes
`cd_programa`	str	Yes	—	Yes
`cd_sub_programa`	str	Yes	—	Yes
`fonte_recurso`	str	Yes	—	Yes
`cd_fonte_recurso`	str	Yes	—	Yes
`tipo_seguro`	str	Yes	—	Yes
`cd_tipo_seguro`	str	Yes	—	Yes
`modalidade`	str	Yes	—	Yes
`cd_modalidade`	str	Yes	—	Yes
`atividade`	str	Yes	—	Yes
`cd_atividade`	str	Yes	—	Yes
`qtd_contratos`	int	Yes	—	Yes
`valor`	float	Yes	BRL	Yes
`area_financiada`	float	Yes	ha	Yes
`fonte`	str	No	—	Yes

Primary key: [ano_emissao, mes_emissao, uf, produto, finalidade, cd_programa, cd_sub_programa, cd_fonte_recurso, cd_tipo_seguro, cd_atividade, cd_modalidade]

Constraints: ano_emissao >= 2013, 1 <= mes_emissao <= 12, qtd_contratos >= 0, valor >= 0, area_financiada >= 0

The key is the entities' grain. agrobr's $select is the whole entity, with every field in the OData $metadata. The 11 dimensions do not repeat across 3,462 records: 12 queries of crop year 2024/25 by product and state, and 2 months of 2024 without a state filter (soybean operating costs in October, with 18 states, and cattle investment in March, with 25). regiao depends on the state and safra on the month, so they stay out of the key.
Names from BCB's domain tables:
programa: the part of the official description before the first " - ", as in agregacao="programa";
tipo_seguro: the official description;
fonte_recurso, modalidade and atividade: the whole official description. For funding sources, the part before " - " would merge distinct codes: 4 sources would become "POUPANÇA RURAL", and 3, "LETRA DE CRÉDITO DO AGRONEGÓCIO (LCA)";
code outside the table: fonte_recurso, modalidade and atividade stay null, without a guess; programa and tipo_seguro come out as Desconhecido (<code>), as in the other aggregations.
Codes as the source publishes them. cd_modalidade comes out as "01", and the official table uses "1": the name is resolved by the number. The purpose × activity × modality combination of the 2,508 records of the 11 queries exists in the official table.
Sub-programme: code only, as in 1.1.0. The name is in BCB's Subprograma domain table and depends on the programme.
Area: through OData, area_financiada is null. Operating costs publish an empty AreaCusteio, and investment and marketing have no area.
Sum: summed by crop year, state, product and purpose, the records equal agregacao="uf" of the same call.
Current crop year: the same warning as the other aggregations, in MetaInfo.validation_warnings and in source_details["safra_em_curso"].
pandas types: those of the empty contract (str → the installed pandas default text dtype, str on pandas 3 and object on 2; int → Int64; float → float64), including the empty result. With as_polars=True: Utf8, Int64 and Float64.
MetaInfo: schema_version and contract_version are 1.0, and source_details["contract"] is bcb.credito_rural_registro.

Version history¶

Version	Change
v1.0	Initial contract (2.0.0)

Example¶

from agrobr import bcb, datasets

df = await bcb.credito_rural("soja", safra="2024/25", uf="MT", agregacao="registro")
by_source = df.groupby(["mes_emissao", "fonte_recurso"])["valor"].sum()

from_dataset = await datasets.credito_rural("milho", safra="2024/25", agregacao="registro")

JSON Schema¶

Available at agrobr/schemas/bcb_credito_rural_registro.json.

from agrobr.contracts import get_contract

contract = get_contract("bcb_credito_rural_registro")
print(contract.to_json())