PYield 0.35.2__tar.gz → 0.35.4__tar.gz

{pyield-0.35.2 → pyield-0.35.4}/.gitignore

@@ -21,10 +21,6 @@ share/python-wheels/
 *.egg
 MANIFEST
 
-# pdm
-pdm.lock
-.pdm-python
-
 # uv
 uv.lock
 
@@ -34,8 +30,6 @@ uv.lock
 env/
 venv/
 ENV/
-env.bak/
-venv.bak/
 
 # Jupiter Notebook
 .ipynb_checkpoints

{pyield-0.35.2 → pyield-0.35.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PYield
-Version: 0.35.2
+Version: 0.35.4
 Summary: A Python library for analysis of fixed income instruments in Brazil
 Project-URL: Source, https://github.com/crdcj/PYield
 Author-email: Carlos Carvalho <cr.cj@outlook.com>

pyield-0.35.4/pyield/__about__.py

@@ -0,0 +1 @@
+__version__ = "0.35.4"

{pyield-0.35.2 → pyield-0.35.4}/pyield/__init__.py

@@ -1,29 +1,30 @@
 import logging
 
-from pyield import anbima, b3, bc, bday, ibge
+from pyield import anbima, b3, bc, bday, ibge, tn
 from pyield.__about__ import __version__
 from pyield.b3 import di1, futures
 from pyield.fwd import forward, forwards
 from pyield.interpolator import Interpolator
-from pyield.tpf import lft, ltn, ntnb, ntnc, ntnf
+from pyield.tn import lft, ltn, ntnb, ntnc, ntnf
 
 __all__ = [
     "__version__",
     "anbima",
+    "bc",
+    "b3",
+    "ibge",
+    "tn",
     "bday",
     "di1",
     "forwards",
     "forward",
     "futures",
-    "ibge",
     "Interpolator",
     "lft",
     "ltn",
     "ntnb",
     "ntnc",
     "ntnf",
-    "bc",
-    "b3",
 ]
 
 # Configura o logger do pacote principal com um NullHandler

{pyield-0.35.2 → pyield-0.35.4}/pyield/b3/di1.py

@@ -61,7 +61,7 @@ def data(
     # Remove duplicates and sort dates
     dates = dates.drop_duplicates().sort_values().reset_index(drop=True)
 
-    df = get_cached_dataset("DI").query("TradeDate in @dates").reset_index(drop=True)
+    df = get_cached_dataset("di1").query("TradeDate in @dates").reset_index(drop=True)
 
     today = dt.now(TIMEZONE_BZ)
     last_bday = bday.last_business_day()
@@ -383,7 +383,7 @@ def eod_dates() -> pd.Series:
         dtype: datetime64[ns]
     """
     available_dates = (
-        get_cached_dataset("DI")
+        get_cached_dataset("di1")
         .drop_duplicates(subset=["TradeDate"])["TradeDate"]
         .sort_values(ascending=True)
         .reset_index(drop=True)

{pyield-0.35.2 → pyield-0.35.4}/pyield/bc/auction.py

@@ -16,8 +16,8 @@ from pyield import bday
 from pyield import date_converter as dc
 from pyield.date_converter import DateScalar
 from pyield.retry import default_retry
-from pyield.tpf.ntnb import duration as duration_b
-from pyield.tpf.ntnf import duration as duration_f
+from pyield.tn.ntnb import duration as duration_b
+from pyield.tn.ntnf import duration as duration_f
 
 """
 Dicionário com o mapeamento das colunas da API do BC para o DataFrame final
@@ -72,18 +72,17 @@ def _pre_process_df(df: pd.DataFrame) -> pd.DataFrame:
 
 
 def _process_df(df: pd.DataFrame) -> pd.DataFrame:
-    """
-    A API retorna somente os valores de SR e FR
-    Assim, vamos ter que somar SR com FR para obter os valores totais
-    Por isso, vamos remover valores nulos de SR e FR para o nulo não ser propagado
-    """
+    # A API retorna somente os valores de SR e FR
+    # Assim, vamos ter que somar SR com FR para obter os valores totais
+
+    # Remover valores nulos de SR e FR para o nulo não ser propagado
     df["AcceptedQuantitySR"] = df["AcceptedQuantitySR"].fillna(0)
     df["OfferedQuantitySR"] = df["OfferedQuantitySR"].fillna(0)
 
     df["OfferedQuantity"] = df["OfferedQuantityFR"] + df["OfferedQuantitySR"]
     df["AcceptedQuantity"] = df["AcceptedQuantityFR"] + df["AcceptedQuantitySR"]
-    # Total value of the auction in R$ millions -> convert it to unit value
-    df["Value"] = (1_000_000 * df["Value"]).round(0).astype("Int64")
+    # Total value of the auction in R$ millions -> convert to R$
+    df["Value"] = (1_000_000 * df["Value"]).round(2)
 
     # Remove the percentage sign and round to 6 decimal places (4 decimal places in %)
     # Before 25/08/2015, there were no rounding rules for the rates in the BC API
@@ -96,12 +95,12 @@ def _process_df(df: pd.DataFrame) -> pd.DataFrame:
     first_round_ratio = first_round_ratio.where(df["AcceptedQuantityFR"] != 0, 0)
     # O dado do financeiro do BC está em milhões com uma casa decimal de precisão
     # Portanto, podemos converter para inteiro sem perda de informação
-    df["ValueFR"] = (first_round_ratio * df["Value"]).round(0).astype("Int64")
+    df["ValueFR"] = (first_round_ratio * df["Value"]).round(2)
 
     # Calculate the financial value of the second round
     second_round_ratio = df["AcceptedQuantitySR"] / df["AcceptedQuantity"]
     second_round_ratio = second_round_ratio.where(df["AcceptedQuantitySR"] != 0, 0)
-    df["ValueSR"] = (second_round_ratio * df["Value"]).round(0).astype("Int64")
+    df["ValueSR"] = (second_round_ratio * df["Value"]).round(2)
 
     bond_mappping = {
         100000: "LTN",
@@ -114,9 +113,9 @@ def _process_df(df: pd.DataFrame) -> pd.DataFrame:
 
     # Em 11/06/2024 o BC passou a informar nas colunas de cotacao os valores dos PUs
     # Isso afeta somente os títulos LFT e NTN-B
-    change_Date = pd.Timestamp("2024-06-11")
+    change_date = pd.Timestamp("2024-06-11")
     adjusted_price = (df["ValueFR"] / df["AcceptedQuantityFR"]).round(6)
-    is_after_change_Date = df["Date"] >= change_Date
+    is_after_change_Date = df["Date"] >= change_date
     is_ltn_or_ntnf = df["BondType"].isin(["LTN", "NTN-F"])
     keep_avg_price = is_after_change_Date | is_ltn_or_ntnf
     df["AvgPrice"] = df["AvgPrice"].where(keep_avg_price, adjusted_price)
@@ -128,63 +127,69 @@ def _process_df(df: pd.DataFrame) -> pd.DataFrame:
 
 
 def _adjust_null_values(df: pd.DataFrame) -> pd.DataFrame:
-    # Onde não há quantidade aceita, não há nem taxa e nem PU definidos.
+    # Onde não há quantidade aceita na primeira volta, não há taxa ou PU definidos.
     is_accepted = df["AcceptedQuantityFR"] != 0
-    df["AvgRate"] = df["AvgRate"].where(is_accepted, pd.NA)
-    df["CutRate"] = df["CutRate"].where(is_accepted, pd.NA)
-    df["AvgPrice"] = df["AvgPrice"].where(is_accepted, pd.NA)
-    df["CutPrice"] = df["CutPrice"].where(is_accepted, pd.NA)
+    cols_to_update = ["AvgRate", "CutRate", "AvgPrice", "CutPrice"]
+    df[cols_to_update] = df[cols_to_update].where(is_accepted, pd.NA)
 
     return df
 
 
-def _add_dv01(df: pd.DataFrame) -> pd.DataFrame:
-    # DV01 irá armazenar o valor do DV01 por título até o final da função
-    # Ao final, o DV01 será multiplicado pela quantidade aceita
-    df["DV01"] = 0
-    is_accepted = df["AcceptedQuantity"] != 0  # noqa
-    df_not_accepted = df.query("~@is_accepted").reset_index(drop=True)
-    df_accepted = df.query("@is_accepted").reset_index(drop=True)
+def _add_duration(df: pd.DataFrame) -> pd.DataFrame:
+    df["Duration"] = 0.0
+    df_lft = df.query("BondType in ['LFT']").reset_index(drop=True)
 
-    df_lft = df_accepted.query("BondType == 'LFT'").reset_index(drop=True)
-
-    df_ltn = df_accepted.query("BondType == 'LTN'").reset_index(drop=True)
+    df_ltn = df.query("BondType == 'LTN'").reset_index(drop=True)
     if not df_ltn.empty:
         df_ltn["Duration"] = df_ltn["BDToMat"] / 252
-        mduration = df_ltn["Duration"] / (1 + df_ltn["AvgRate"])
-        df_ltn["DV01"] = 0.0001 * mduration * df_ltn["AvgPrice"]
-
-    def compute_f_duration(row):
-        return duration_f(row["Date"], row["Maturity"], row["CutRate"])
 
-    df_ntnf = df_accepted.query("BondType == 'NTN-F'").reset_index(drop=True)
+    df_ntnf = df.query("BondType == 'NTN-F'").reset_index(drop=True)
     if not df_ntnf.empty:
-        df_ntnf["Duration"] = df_ntnf.apply(compute_f_duration, axis=1)
-        df_ntnf["Duration"] = df_ntnf["Duration"].astype("Float64")
-        mduration = df_ntnf["Duration"] / (1 + df_ntnf["AvgRate"])
-        df_ntnf["DV01"] = 0.0001 * mduration * df_ntnf["AvgPrice"]
-
-    def compute_b_duration(row):
-        return duration_b(row["Date"], row["Maturity"], row["CutRate"])
+        df_ntnf["Duration"] = df_ntnf.apply(
+            lambda row: duration_f(row["Settlement"], row["Maturity"], row["AvgRate"]),
+            axis=1,
+        )
 
-    df_ntnb = df_accepted.query("BondType == 'NTN-B'").reset_index(drop=True)
+    df_ntnb = df.query("BondType == 'NTN-B'").reset_index(drop=True)
     if not df_ntnb.empty:
-        df_ntnb["Duration"] = df_ntnb.apply(compute_b_duration, axis=1)
-        df_ntnb["Duration"] = df_ntnb["Duration"].astype("Float64")
-        mduration = df_ntnb["Duration"] / (1 + df_ntnb["AvgRate"])
-        df_ntnb["DV01"] = 0.0001 * mduration * df_ntnb["AvgPrice"]
+        df_ntnb["Duration"] = df_ntnb.apply(
+            lambda row: duration_b(row["Settlement"], row["Maturity"], row["AvgRate"]),
+            axis=1,
+        )
 
-    df = pd.concat([df_not_accepted, df_lft, df_ltn, df_ntnf, df_ntnb])
+    df = pd.concat([df_lft, df_ltn, df_ntnf, df_ntnb]).reset_index(drop=True)
+    df["Duration"] = df["Duration"].astype("Float64")
 
-    df["DV01FR"] = df["DV01"] * df["AcceptedQuantityFR"]
-    df["DV01FR"] = df["DV01FR"].round(0).astype("Int64")
+    return df
+
+
+def _add_dv01(df: pd.DataFrame) -> pd.DataFrame:
+    # DV01 por título calculado com base nos valores da primeira rodada
+    mduration = df["Duration"] / (1 + df["AvgRate"])
+    dv01 = 0.0001 * mduration * df["AvgPrice"]
 
-    df["DV01SR"] = df["DV01"] * df["AcceptedQuantitySR"]
-    df["DV01SR"] = df["DV01SR"].round(0).astype("Int64")
+    # Valores totais de DV01 para o leilão
+    df["DV01"] = dv01 * df["AcceptedQuantity"]
+    df["DV01FR"] = dv01 * df["AcceptedQuantityFR"]
+    df["DV01SR"] = dv01 * df["AcceptedQuantitySR"]
 
-    # DV01 irá armazenar o valor total do leilão
-    df["DV01"] *= df["AcceptedQuantity"]
-    df["DV01"] = df["DV01"].round(0).astype("Int64")
+    for col in ["DV01", "DV01FR", "DV01SR"]:
+        # Definiri DV01 nulos do leilão como 0
+        df[col] = df[col].fillna(0).round(2)
+
+    # Forçar 0 nas LFT, pois não há DV01
+    df.loc[df["BondType"] == "LFT", ["DV01", "DV01FR", "DV01SR"]] = 0.0
+
+    return df
+
+
+def _add_avg_maturity(df: pd.DataFrame) -> pd.DataFrame:
+    # Na metodolgia do Tesouro Nacional, a maturidade média é a mesma que a duração
+    df["AvgMaturity"] = df["Duration"]
+
+    # Para LFT, a maturidade média é calculada como o n. de dias úteis até o vencimento
+    is_lft = df["BondType"] == "LFT"
+    df.loc[is_lft, "AvgMaturity"] = df.loc[is_lft, "BDToMat"] / 252
 
     return df
 
@@ -197,10 +202,11 @@ def _sort_and_reorder_columns(df: pd.DataFrame) -> pd.DataFrame:
         "Ordinance",
         "Buyer",
         "BondType",
+        "SelicCode",
         "Maturity",
         "BDToMat",
         "Duration",
-        "SelicCode",
+        "AvgMaturity",
         "Value",
         "ValueFR",
         "ValueSR",
@@ -232,7 +238,9 @@ def _fetch_df_from_url(url: str) -> pd.DataFrame:
         df = _pre_process_df(df)
         df = _adjust_null_values(df)
         df = _process_df(df)
+        df = _add_duration(df)
         df = _add_dv01(df)
+        df = _add_avg_maturity(df)
         df = _sort_and_reorder_columns(df)
         return df
     except Exception:
@@ -243,7 +251,7 @@ def _fetch_df_from_url(url: str) -> pd.DataFrame:
 def auctions(
     start: DateScalar | None = None,
     end: DateScalar | None = None,
-    auction_type: Literal["Sell", "Buy"] | None = None,
+    auction_type: Literal["sell", "buy"] | None = None,
 ) -> pd.DataFrame:
     """
     Recupera dados de leilões para um determinado período e tipo de leilão da API do BC.
@@ -286,7 +294,7 @@ def auctions(
             leilão desde a data mais antiga disponível até a data de `end`.
             Se `start` e `end` forem `None`, a série histórica completa será retornada.
             Padrão é `None`.
-        auction_type (Literal["Sell", "Buy"], opcional): O tipo de leilão para filtrar
+        auction_type (Literal["sell", "buy"], opcional): O tipo de leilão para filtrar
             diretamente na API. Padrão é `None` (retorna todos os tipos de leilão).
 
     Returns:
@@ -305,8 +313,12 @@ def auctions(
             - Ordinance: Edital normativo associado ao leilão.
             - Buyer: Categoria do comprador (ex: "TodoMercado", "SomenteDealerApto").
             - BondType: Categoria do título (ex: "LTN", "LFT", "NTN-B", "NTN-F").
-            - Maturity: Data de vencimento do título.
             - SelicCode: Código do título no sistema Selic.
+            - Maturity: Data de vencimento do título.
+            - DV01: Valor do DV01 total do leilão em R$.
+            - DV01FR: DV01 da Primeira Rodada (FR) em R$.
+            - DV01SR: DV01 da Segunda Rodada (SR) em R$.
+            - AvgMaturity: Maturidade média do título (em anos).
             - Value: Valor total do leilão em R$ (FR + SR).
             - ValueFR: Valor da primeira rodada (FR) do leilão em R$.
             - ValueSR: Valor da segunda rodada (SR) em R$.
@@ -322,9 +334,6 @@ def auctions(
             - CutRate: Taxa de corte.
             - BDToMat: Dias úteis até o vencimento.
             - Duration: Duration (Duração) do título.
-            - DV01: Valor do DV01 total do leilão em R$.
-            - DV01FR: DV01 da Primeira Rodada (FR) em R$.
-            - DV01SR: DV01 da Segunda Rodada (SR) em R$.
     """
     url = BASE_API_URL
     if start:
@@ -338,8 +347,8 @@ def auctions(
         url += f"&@dataMovimentoFim='{end_str}'"
 
     # Mapeamento do auction_type para o valor esperado pela API
-    auction_type_mapping = {"Sell": "Venda", "Buy": "Compra"}
-    auction_type_api_value = auction_type_mapping.get(str(auction_type))
+    auction_type_mapping = {"sell": "Venda", "buy": "Compra"}
+    auction_type_api_value = auction_type_mapping.get(str(auction_type.lower()))
     # Adiciona o parâmetro tipoOferta à URL se auction_type for fornecido
     if auction_type_api_value:
         url += f"&@tipoOferta='{auction_type_api_value}'"

{pyield-0.35.2 → pyield-0.35.4}/pyield/bc/trades_intraday.py

@@ -1,3 +1,8 @@
+"""
+Fetches real-time secondary trading data for domestic Federal Public Debt (FPD)
+https://www.bcb.gov.br/htms/selic/selicprecos.asp?frame=1
+"""
+
 import datetime as dt
 import io
 import logging
@@ -18,7 +23,13 @@ test_file = Path(__file__).parent.parent.parent / "tests/data/negocios-registrad
 
 
 def _fetch_csv_from_url() -> str:
-    FILE_URL = "https://www3.bcb.gov.br/novoselic/NegociosRegistradosDownload"
+    """
+    Example URL for the CSV file containing intraday trades data:
+        https://www3.bcb.gov.br/novoselic/rest/precosNegociacao/pub/download/estatisticas/02-06-2025
+    """
+    today = dt.datetime.now(BZ_TIMEZONE).date()
+    formatted_date = today.strftime("%d-%m-%Y")
+    FILE_URL = f"https://www3.bcb.gov.br/novoselic/rest/precosNegociacao/pub/download/estatisticas/{formatted_date}"
     r = requests.get(FILE_URL, timeout=30)  # API usually takes 10s to respond
     r.raise_for_status()
     r.encoding = "iso-8859-15"

{pyield-0.35.2 → pyield-0.35.4}/pyield/data_cache.py

@@ -1,6 +1,7 @@
 import functools
 import logging
 from datetime import datetime
+from typing import Literal
 from zoneinfo import ZoneInfo
 
 import pandas as pd
@@ -16,12 +17,12 @@ BASE_DATA_URL = f"{GIT_URL}/data"
 
 # Dicionário global para configurações dos datasets usando 'filename'
 DATASET_CONFIGS = {
-    "DI": {
+    "di1": {
         "filename": "b3_di.parquet",  # Apenas o nome do arquivo
         "date_column": "TradeDate",
         "name": "Futuro de DI (B3)",
     },
-    "TPF": {
+    "tpf": {
         "filename": "anbima_tpf.parquet",  # Apenas o nome do arquivo
         "date_column": "ReferenceDate",
         "name": "TPF (ANBIMA)",
@@ -73,6 +74,6 @@ def _get_dataset_with_ttl(dataset_id: str, date_key: str) -> pd.DataFrame:
         raise  # Re-raise a exceção para que o chamador saiba que falhou
 
 
-def get_cached_dataset(dataset_id: str) -> pd.DataFrame:
+def get_cached_dataset(dataset_id: Literal["di1", "tpf"]) -> pd.DataFrame:
     """Obtém um dataset configurado pelo seu ID. O cache expira diariamente."""
-    return _get_dataset_with_ttl(dataset_id, _get_today_date_key())
+    return _get_dataset_with_ttl(dataset_id.lower(), _get_today_date_key())

pyield-0.35.4/pyield/tn/__init__.py

@@ -0,0 +1,3 @@
+from pyield.tn.auction import benchmarks
+
+__all__ = ["benchmarks"]

pyield-0.35.4/pyield/tn/auction.py

@@ -0,0 +1,91 @@
+"""
+Manual da API: https://portal-conhecimento.tesouro.gov.br/catalogo-componentes/api-leil%C3%B5es
+Financeiro é sempre o financeiro aceito.
+Os dados do BCB só existem na API do TN.
+"""
+
+import logging
+
+import pandas as pd
+import requests
+
+BENCHMARKS_COLUMN_MAPPING = {
+    "BENCHMARK": "Benchmark",
+    "VENCIMENTO": "MaturityDate",
+    "TÍTULO": "BondType",
+    "INÍCIO": "StartDate",
+    "TERMINO": "EndDate",
+}
+
+
+API_URL = "https://apiapex.tesouro.gov.br/aria/v1/api-leiloes-pub/custom/benchmarks?incluir_historico="
+
+
+def benchmarks(include_history: bool = False) -> pd.DataFrame:
+    """Fetches benchmark data for Brazilian Treasury Bonds from the TN API.
+
+    This function retrieves current or historical benchmark data for various Brazilian
+    Treasury bond types (e.g., LTN, LFT, NTN-B). The data is sourced directly from the
+    official Tesouro Nacional API.
+
+    Args:
+        include_history (bool, optional): If `True`, includes historical benchmark data.
+            If `False` (default), only current benchmarks are returned.
+
+    Returns:
+        pd.DataFrame: A pandas DataFrame containing the benchmark data.
+            The DataFrame includes the following columns:
+
+            *   `Benchmark` (str): The name or identifier of the benchmark
+                (e.g., 'LFT 3 anos').
+            *   `MaturityDate` (datetime64[ns]): The maturity date of the benchmark.
+            *   `BondType` (str): The type of the bond (e.g., 'LTN', 'LFT', 'NTN-B').
+            *   `StartDate` (datetime64[ns]): The start date for the benchmark's period.
+            *   `EndDate` (datetime64[ns]): The end date for the benchmark's period.
+
+    Notes:
+        *   Data is sourced from the official Tesouro Nacional (Brazilian Treasury) API.
+        *   An retry mechanism is implemented for SSL certificate verification errors.
+        *   The API documentation can be found at:
+            https://portal-conhecimento.tesouro.gov.br/catalogo-componentes/api-leil%C3%B5es
+        *   Rows with any `NaN` values are dropped before returning the DataFrame.
+
+    Examples:
+        >>> # Get current benchmarks
+        >>> from pyield import tn
+        >>> df_current = tn.benchmarks()
+
+        >>> # Get historical benchmarks
+        >>> df_history = benchmarks(include_history=True)
+        >>> df_history.head()
+            Benchmark MaturityDate BondType  StartDate    EndDate
+        0  LFT 6 anos   2020-03-01      LFT 2014-01-01 2014-06-30
+        1  LFT 6 anos   2020-09-01      LFT 2014-07-01 2014-12-31
+        2  LFT 6 anos   2021-03-01      LFT 2015-01-01 2015-04-30
+        3  LFT 6 anos   2021-09-01      LFT 2015-05-01 2015-12-31
+        4   LFT 1 ano   2022-03-01      LFT 2020-10-15 2021-03-31
+    """
+    session = requests.Session()
+    include_history_param = "S" if include_history else "N"
+    api_endpoint = API_URL + include_history_param
+    try:
+        stn_benchmarks = session.get(api_endpoint)
+    except requests.exceptions.SSLError as e:
+        logging.error(
+            f"SSL error encountered: {e}. Retrying without certificate verification."
+        )
+        stn_benchmarks = session.get(api_endpoint, verify=False)
+    response_dict = stn_benchmarks.json()
+    df = pd.DataFrame(response_dict["registros"]).dropna()
+    df = df.convert_dtypes()
+    # A API pode retornar vazio, então é necessário verificar
+    if not df.empty:
+        df["VENCIMENTO"] = pd.to_datetime(df["VENCIMENTO"])
+        df["TERMINO"] = pd.to_datetime(df["TERMINO"])
+        df["INÍCIO"] = pd.to_datetime(df["INÍCIO"])
+        df["BENCHMARK"] = df["BENCHMARK"].str.strip()
+        df["TÍTULO"] = df["TÍTULO"].str.strip()
+    else:
+        logging.warning("Não foi possível obter os benchmarks do STN.")
+
+    return df.rename(columns=BENCHMARKS_COLUMN_MAPPING)

{pyield-0.35.2/pyield/tpf → pyield-0.35.4/pyield/tn}/lft.py

@@ -3,7 +3,7 @@ import pandas as pd
 from pyield import anbima, bday
 from pyield import date_converter as dc
 from pyield.date_converter import DateScalar
-from pyield.tpf import tools
+from pyield.tn import tools
 
 
 def data(date: DateScalar) -> pd.DataFrame:

{pyield-0.35.2/pyield/tpf → pyield-0.35.4/pyield/tn}/ltn.py

@@ -4,7 +4,7 @@ from pyield import anbima, bday
 from pyield import date_converter as dc
 from pyield.b3 import di1
 from pyield.date_converter import DateScalar
-from pyield.tpf import tools
+from pyield.tn import tools
 
 FACE_VALUE = 1000
 

{pyield-0.35.2/pyield/tpf → pyield-0.35.4/pyield/tn}/ntnb.py

@@ -3,7 +3,7 @@ import pandas as pd
 
 import pyield.date_converter as dc
 import pyield.interpolator as ip
-import pyield.tpf.tools as bt
+import pyield.tn.tools as bt
 from pyield import anbima, bday
 from pyield.date_converter import DateScalar
 
@@ -541,14 +541,20 @@ def duration(
         >>> ntnb.duration("23-08-2024", "15-08-2060", 0.061005)
         15.083054313130464
     """
+    # Return NaN if any input is NaN
+    if any(pd.isna(x) for x in [settlement, maturity, rate]):
+        return float("NaN")
+
+    # Validate and normalize dates
     settlement = dc.convert_input_dates(settlement)
     maturity = dc.convert_input_dates(maturity)
 
     df = cash_flows(settlement, maturity)
     df["BY"] = bday.count(settlement, df["PaymentDate"]) / 252
     df["DCF"] = df["CashFlow"] / (1 + rate) ** df["BY"]
-    np_duration = (df["DCF"] * df["BY"]).sum() / df["DCF"].sum()
-    return float(np_duration)
+    duration = (df["DCF"] * df["BY"]).sum() / df["DCF"].sum()
+    # Return the duration as native float
+    return float(duration)
 
 
 def dv01(

{pyield-0.35.2/pyield/tpf → pyield-0.35.4/pyield/tn}/ntnc.py

@@ -2,7 +2,7 @@ import numpy as np
 import pandas as pd
 
 import pyield.date_converter as dc
-import pyield.tpf.tools as bt
+import pyield.tn.tools as bt
 from pyield import anbima, bday
 from pyield.date_converter import DateScalar
 
@@ -266,11 +266,17 @@ def duration(
         >>> ntnc.duration("21-03-2025", "01-01-2031", 0.067626)
         4.405363320448003
     """
+    # Return NaN if any input is NaN
+    if any(pd.isna(x) for x in [settlement, maturity, rate]):
+        return float("NaN")
+
+    # Validate and normalize dates
     settlement = dc.convert_input_dates(settlement)
     maturity = dc.convert_input_dates(maturity)
 
     df = cash_flows(settlement, maturity)
     df["BY"] = bday.count(settlement, df["PaymentDate"]) / 252
     df["DCF"] = df["CashFlow"] / (1 + rate) ** df["BY"]
-    np_duration = (df["DCF"] * df["BY"]).sum() / df["DCF"].sum()
-    return float(np_duration)
+    duration = (df["DCF"] * df["BY"]).sum() / df["DCF"].sum()
+    # Return the duration as native float
+    return float(duration)

{pyield-0.35.2/pyield/tpf → pyield-0.35.4/pyield/tn}/ntnf.py

@@ -8,7 +8,7 @@ from pyield import date_converter as dc
 from pyield import interpolator as ip
 from pyield.b3 import di1
 from pyield.date_converter import DateScalar
-from pyield.tpf import tools
+from pyield.tn import tools
 
 """
 Constants calculated as per Anbima Rules
@@ -645,14 +645,20 @@ def duration(
         >>> ntnf.duration("02-09-2024", "01-01-2035", 0.121785)
         6.32854218039796
     """
+    # Return NaN if any input is NaN
+    if any(pd.isna(x) for x in [settlement, maturity, rate]):
+        return float("NaN")
+
+    # Validate and normalize input dates
     settlement = dc.convert_input_dates(settlement)
     maturity = dc.convert_input_dates(maturity)
 
     df = cash_flows(settlement, maturity)
     df["BY"] = bday.count(settlement, df["PaymentDate"]) / 252
     df["DCF"] = df["CashFlow"] / (1 + rate) ** df["BY"]
-    np_duration = (df["DCF"] * df["BY"]).sum() / df["DCF"].sum()
-    return float(np_duration)
+    duration = (df["DCF"] * df["BY"]).sum() / df["DCF"].sum()
+    # Return the duration as native float
+    return float(duration)
 
 
 def dv01(

pyield-0.35.2/pyield/__about__.py

@@ -1 +0,0 @@
-__version__ = "0.35.2"