PYield 0.30.0__tar.gz → 0.30.2__tar.gz

{pyield-0.30.0 → pyield-0.30.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PYield
-Version: 0.30.0
+Version: 0.30.2
 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>
@@ -34,7 +34,7 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Topic :: Office/Business :: Financial :: Investment
 Requires-Python: >=3.11
 Requires-Dist: numpy>=2.0.0
-Requires-Dist: pandas[excel,html,parquet]>=2.0.0
+Requires-Dist: pandas[html,parquet]>=2.0.0
 Requires-Dist: requests>=2.31.0
 Requires-Dist: tenacity>=9.0.0
 Description-Content-Type: text/markdown

pyield-0.30.2/pyield/__about__.py

@@ -0,0 +1 @@
+__version__ = "0.30.2"

{pyield-0.30.0 → pyield-0.30.2}/pyield/anbima/__init__.py

@@ -1,3 +1,4 @@
+from pyield.anbima.ettj import last_ettj
 from pyield.anbima.ima import last_ima
 from pyield.anbima.imaq import imaq
 from pyield.anbima.ipca import ipca_projection
@@ -10,4 +11,5 @@ __all__ = [
     "tpf_data",
     "tpf_rates",
     "tpf_pre_maturities",
+    "last_ettj",
 ]

pyield-0.30.2/pyield/anbima/ettj.py

@@ -0,0 +1,135 @@
+import logging
+from io import StringIO
+
+import pandas as pd
+import requests
+
+from pyield.forwards import forward_rates
+from pyield.retry import default_retry
+
+logger = logging.getLogger(__name__)
+LAST_ETTJ_URL = "https://www.anbima.com.br/informacoes/est-termo/CZ-down.asp"
+
+# Anbima ETTJ data has 4 decimal places in percentage values
+# We will round to 6 decimal places to avoid floating point errors
+ROUND_DIGITS = 6
+
+
+@default_retry
+def _get_last_content() -> str:
+    """Fetches the raw yield curve data from ANBIMA."""
+    request_payload = {
+        "Idioma": "PT",
+        "Dt_Ref": "",
+        "saida": "csv",
+    }
+    r = requests.post(LAST_ETTJ_URL, data=request_payload)
+    r.raise_for_status()
+    return r.text
+
+
+def _convert_text_to_df(text: str) -> pd.DataFrame:
+    """Converts the raw yield curve text to a DataFrame."""
+    raw_df = pd.read_csv(
+        StringIO(text),
+        sep=";",
+        encoding="latin1",
+        skiprows=5,
+        dtype_backend="numpy_nullable",
+    )
+
+    idx = raw_df.query("Vertices =='PREFIXADOS (CIRCULAR 3.361)'").first_valid_index()  # noqa
+    raw_df.query("index < @idx", inplace=True)
+    csv_buffer = StringIO()
+
+    # Save the DataFrame as CSV in memory buffer to be read again
+    # It's easier than handling values with commas and dots directly in the DataFrame
+    raw_df.to_csv(csv_buffer, index=False)
+    csv_buffer.seek(0)  # Move the buffer cursor to the beginning
+
+    df = pd.read_csv(
+        csv_buffer,
+        decimal=",",
+        thousands=".",
+        dtype_backend="numpy_nullable",
+    )
+    file_date_str = text[0:10]  # 09/09/2024
+    file_date = pd.to_datetime(file_date_str, format="%d/%m/%Y")
+    df["date"] = file_date
+
+    return df
+
+
+def _process_df(df: pd.DataFrame) -> pd.DataFrame:
+    """Processes the raw yield curve DataFrame to calculate rates and forward rates."""
+    # Rename columns
+    rename_dict = {
+        "Vertices": "vertex",
+        "ETTJ IPCA": "real_rate",
+        "ETTJ PREF": "nominal_rate",
+        "Inflação Implícita": "implied_inflation",
+    }
+    df.rename(columns=rename_dict, inplace=True)
+
+    # Divide float columns by 100 and round to 6 decimal places
+    df["real_rate"] = (df["real_rate"] / 100).round(ROUND_DIGITS)
+    df["nominal_rate"] = (df["nominal_rate"] / 100).round(ROUND_DIGITS)
+    df["implied_inflation"] = (df["implied_inflation"] / 100).round(ROUND_DIGITS)
+
+    df["real_rate_fwd"] = forward_rates(
+        business_days=df["vertex"],
+        zero_rates=df["real_rate"],
+        groupby_dates=df["date"],
+    )
+    df["nominal_rate_fwd"] = forward_rates(
+        business_days=df["vertex"],
+        zero_rates=df["nominal_rate"],
+        groupby_dates=df["date"],
+    )
+    df["implied_inflation_fwd"] = forward_rates(
+        business_days=df["vertex"],
+        zero_rates=df["implied_inflation"],
+        groupby_dates=df["date"],
+    )
+
+    column_order = [
+        "date",
+        "vertex",
+        "nominal_rate",
+        "real_rate",
+        "implied_inflation",
+        "nominal_rate_fwd",
+        "real_rate_fwd",
+        "implied_inflation_fwd",
+    ]
+    df = df[column_order].copy()
+
+    return df
+
+
+def last_ettj() -> pd.DataFrame:
+    """
+    Retrieves and processes the latest Brazilian yield curve data from ANBIMA.
+
+    This function fetches the most recent yield curve data published by ANBIMA,
+    containing real rates (IPCA-indexed), nominal rates, and implied inflation
+    at various vertices (time points). Both spot rates and forward rates are
+    calculated and included in the returned DataFrame.
+
+    Returns:
+        pd.DataFrame: A DataFrame containing the following columns:
+            - date: Reference date of the yield curve
+            - vertex: Time point in business days
+            - nominal_rate: Zero-coupon nominal interest rate
+            - real_rate: Zero-coupon real interest rate (IPCA-indexed)
+            - implied_inflation: Implied inflation rate (break-even inflation)
+            - nominal_rate_fwd: Forward nominal interest rate
+            - real_rate_fwd: Forward real interest rate
+            - implied_inflation_fwd: Forward implied inflation rate
+
+    Note:
+        All rates are expressed in decimal format (e.g., 0.12 for 12%).
+    """
+    text = _get_last_content()
+    df = _convert_text_to_df(text)
+    return _process_df(df)

pyield-0.30.2/pyield/anbima/ipca.py

@@ -0,0 +1,85 @@
+import io
+import locale
+import warnings
+from dataclasses import dataclass
+
+import pandas as pd
+import requests
+from bs4 import XMLParsedAsHTMLWarning
+
+warnings.filterwarnings("ignore", category=XMLParsedAsHTMLWarning)
+
+
+@dataclass
+class IndicatorProjection:
+    reference_period: pd.Period  # Reference month as a pd.Period object
+    projected_value: float  # Projected value
+    last_updated: pd.Timestamp  # Date and time of the last update
+
+
+def ipca_projection() -> IndicatorProjection:
+    """
+    Retrieves the current IPCA projection from the ANBIMA website.
+
+    This function makes an HTTP request to the ANBIMA website, extracts HTML tables
+    containing economic indicators, and specifically processes the IPCA projection data.
+
+    Process:
+    1. Accesses the ANBIMA indicators webpage
+    2. Extracts the third table that contains the IPCA projection
+    3. Locates the row labeled as "IPCA1"
+    4. Extracts the projection value and converts it to decimal format
+    5. Extracts and formats the reference month of the projection
+    6. Extracts the date and time of the last update
+
+    Returns:
+        IndicatorProjection: An object containing:
+            - reference_period (pd.Period): Reference period of the projection
+            - projected_value (float): Projected IPCA value as a decimal number
+            - last_updated (pd.Timestamp): Date and time of the last data update
+
+    Raises:
+        requests.RequestException: If there are connection issues with the ANBIMA site
+        ValueError: If the expected data is not found in the page structure
+        locale.Error: If the 'pt_BR.UTF-8' locale is not available on the system
+
+    Notes:
+        - The function requires internet connection to access the ANBIMA website
+        - The structure of the ANBIMA page may change, which could affect the function
+        - Temporarily changes the locale to pt_BR.UTF-8 to process dates in Portuguese
+    """
+    url = "https://www.anbima.com.br/informacoes/indicadores/"
+    r = requests.get(url)
+    r.encoding = "latin1"
+    dfs = pd.read_html(
+        io.StringIO(r.text),
+        flavor="html5lib",
+        decimal=",",
+        thousands=".",
+        dtype_backend="numpy_nullable",
+    )
+    # The IPCA projection is in the third table
+    df = dfs[2]
+
+    last_update_str = df.iat[0, 0].split("Atualização:")[-1].strip()
+    last_update = pd.to_datetime(last_update_str, format="%d/%m/%Y - %H:%M h")
+
+    ipca_row = df.loc[df[0] == "IPCA1"]
+    ipca_value = ipca_row.iloc[0, 2]
+    ipca_value = pd.to_numeric(ipca_value, errors="coerce")
+    ipca_value = round(float(ipca_value) / 100, 4)
+
+    # Extract and format the reference month
+    ipca_date = ipca_row.iloc[0, 1]
+    ipca_date = str(ipca_date)
+    ipca_date = ipca_date.split("(")[-1].split(")")[0]
+    locale.setlocale(locale.LC_TIME, "pt_BR.UTF-8")
+    ipca_date = pd.to_datetime(ipca_date, format="%b/%y")
+    reference_period = ipca_date.to_period("M")
+    locale.setlocale(locale.LC_TIME, "")  # Reset locale to default
+
+    return IndicatorProjection(
+        last_updated=last_update,
+        reference_period=reference_period,
+        projected_value=ipca_value,
+    )

{pyield-0.30.0 → pyield-0.30.2}/pyield/anbima/tpf.py

@@ -1,19 +1,27 @@
 import logging
+from typing import Literal
 from urllib.error import HTTPError
 
 import pandas as pd
 
 from pyield import bday
-from pyield.config import default_retry
 from pyield.data_cache import get_anbima_dataset
 from pyield.date_converter import DateScalar, convert_input_dates
+from pyield.retry import default_retry
 
+BOND_TYPES = Literal["LTN", "NTN-B", "NTN-C", "NTN-F", "LFT"]
 ANBIMA_URL = "https://www.anbima.com.br/informacoes/merc-sec/arqs"
 # URL example: https://www.anbima.com.br/informacoes/merc-sec/arqs/ms240614.txt
 
 logger = logging.getLogger(__name__)
 
 
+def _bond_type_mapping(bond_type: str) -> str:
+    bond_type = bond_type.upper()
+    bond_type_mapping = {"NTNB": "NTN-B", "NTNC": "NTN-C", "NTNF": "NTN-F"}
+    return bond_type_mapping.get(bond_type, bond_type)
+
+
 def _build_file_url(date: pd.Timestamp) -> str:
     url_date = date.strftime("%y%m%d")
     filename = f"ms{url_date}.txt"
@@ -59,8 +67,9 @@ def _process_raw_df(df_raw: pd.DataFrame) -> pd.DataFrame:
 
     # Remove percentage from rates
     rate_cols = [col for col in df.columns if "Rate" in col]
-    # Data has 6 decimal places, so we round to 7 decimal places
-    df[rate_cols] = (df[rate_cols] / 100).round(7)
+    # Rate columns have percentage values with 4 decimal places in percentage values
+    # We will round to 6 decimal places to avoid floating point errors
+    df[rate_cols] = (df[rate_cols] / 100).round(6)
 
     df["ReferenceDate"] = pd.to_datetime(df["ReferenceDate"], format="%Y%m%d")
     df["MaturityDate"] = pd.to_datetime(df["MaturityDate"], format="%Y%m%d")
@@ -86,9 +95,10 @@ def tpf_data(date: DateScalar, bond_type: str | None = None) -> pd.DataFrame:
             Returns an empty DataFrame if data is not available for the
             specified date (weekends, holidays, or unavailable data).
     """
+    date = convert_input_dates(date)
+    date_log = date.strftime("%d/%m/%Y")
+
     try:
-        date = convert_input_dates(date)
-        date_log = date.strftime("%d/%m/%Y")
         file_url = _build_file_url(date)
         df = _read_raw_df(file_url)
         if df.empty:
@@ -99,8 +109,9 @@ def tpf_data(date: DateScalar, bond_type: str | None = None) -> pd.DataFrame:
             return df
         df = _process_raw_df(df)
         if bond_type:
+            bond_type = _bond_type_mapping(bond_type)
             df = df.query("BondType == @bond_type").reset_index(drop=True)
-        return df
+        return df.sort_values(["BondType", "MaturityDate"]).reset_index(drop=True)
     except HTTPError as e:
         if e.code == 404:  # noqa
             logger.info(
@@ -109,8 +120,8 @@ def tpf_data(date: DateScalar, bond_type: str | None = None) -> pd.DataFrame:
             )
             return pd.DataFrame()
         raise  # Propagate other HTTP errors
-    except Exception as e:
-        logger.error(f"Error fetching TPF data for {date_log}: {e}")
+    except Exception:
+        logger.exception(f"Error fetching TPF data for {date_log}")
         raise
 
 

{pyield-0.30.0 → pyield-0.30.2}/pyield/b3/di_over.py

@@ -1,7 +1,10 @@
 import ftplib
+import logging
 
 from pyield.date_converter import DateScalar, convert_input_dates
 
+logger = logging.getLogger(__name__)
+
 
 def di_over(date: DateScalar) -> float:
     """
@@ -46,14 +49,18 @@ def di_over(date: DateScalar) -> float:
             rate = int(raw_rate) / 10000
             return round(rate, 5)
         else:
-            return "Empty file or no data"
+            logger.error(f"No data found for date {date}")
+            return float("nan")
 
     except ValueError as e:
-        return f"Date format error: {e}"
+        logger.error(f"Date format error: {e}")
+        raise
     except ftplib.error_perm as e:
-        return f"File access error: {e}"
-    except Exception as e:
-        return f"Unexpected error: {e}"
+        logger.error(f"File access error: {e}")
+        raise
+    except Exception:
+        logger.exception("Unexpected error")
+        raise
     finally:
         # Ensure FTP connection is closed
         if ftp:
@@ -61,10 +68,3 @@ def di_over(date: DateScalar) -> float:
                 ftp.quit()
             except:  # noqa
                 pass
-
-
-# Example usage
-if __name__ == "__main__":
-    date = "28/02/2025"
-    rate = di_over(date)
-    print(f"DI Rate for {date}: {rate}")

{pyield-0.30.0 → pyield-0.30.2}/pyield/b3/futures/intraday.py

@@ -9,7 +9,7 @@ from pyield import bday
 TIMEZONE_BZ = ZoneInfo("America/Sao_Paulo")
 
 
-def _fetch_b3_df(future_code: str) -> pd.DataFrame:
+def _fetch_b3_df(contract_code: str) -> pd.DataFrame:
     """
     Fetch the latest data for a given future code from B3 derivatives quotation API.
 
@@ -24,13 +24,13 @@ def _fetch_b3_df(future_code: str) -> pd.DataFrame:
         Exception: An exception is raised if the data fetch operation fails.
     """
 
-    url = f"https://cotacao.b3.com.br/mds/api/v1/DerivativeQuotation/{future_code}"
+    url = f"https://cotacao.b3.com.br/mds/api/v1/DerivativeQuotation/{contract_code}"
 
     try:
         r = requests.get(url, timeout=10)
         r.raise_for_status()  # Check for HTTP request errors
     except requests.exceptions.RequestException:
-        raise Exception(f"Failed to fetch data for {future_code}.") from None
+        raise Exception(f"Failed to fetch data for {contract_code}.") from None
 
     r.encoding = "utf-8"  # Explicitly set response encoding to utf-8 for consistency
 

{pyield-0.30.0 → pyield-0.30.2}/pyield/bc/auction.py

@@ -14,8 +14,8 @@ import requests
 
 from pyield import bday
 from pyield import date_converter as dc
-from pyield.config import default_retry
 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
 

{pyield-0.30.0 → pyield-0.30.2}/pyield/bc/bcdata.py

@@ -20,8 +20,8 @@ from urllib.error import HTTPError
 
 import pandas as pd
 
-from pyield.config import default_retry
 from pyield.date_converter import DateScalar, convert_input_dates
+from pyield.retry import default_retry
 
 logger = logging.getLogger(__name__)
 

{pyield-0.30.0 → pyield-0.30.2}/pyield/bc/repo.py

@@ -13,8 +13,8 @@ import requests
 
 from pyield import bday
 from pyield import date_converter as dc
-from pyield.config import default_retry
 from pyield.date_converter import DateScalar
+from pyield.retry import default_retry
 
 """Dicionário com o mapeamento das colunas da API do BC para o DataFrame final
 Chaves com comentário serão descartadas ao final do processamento

{pyield-0.30.0 → pyield-0.30.2}/pyield/bc/sec.py

@@ -2,8 +2,8 @@ import logging
 
 import pandas as pd
 
-from pyield.config import default_retry
 from pyield.date_converter import DateScalar, convert_input_dates
+from pyield.retry import default_retry
 
 logger = logging.getLogger(__name__)
 

{pyield-0.30.0 → pyield-0.30.2}/pyield/bc/vna.py

@@ -1,7 +1,7 @@
 import requests
 
-from pyield.config import default_retry
 from pyield.date_converter import DateScalar, convert_input_dates
+from pyield.retry import default_retry
 
 
 @default_retry

{pyield-0.30.0 → pyield-0.30.2}/pyield/ibge/ipca.py

@@ -3,8 +3,8 @@ import logging
 import pandas as pd
 import requests
 
-from pyield.config import default_retry
 from pyield.date_converter import DateScalar, convert_input_dates
+from pyield.retry import default_retry
 
 logger = logging.getLogger(__name__)
 IPCA_URL = "https://servicodados.ibge.gov.br/api/v3/agregados/6691/periodos/"

{pyield-0.30.0 → pyield-0.30.2}/pyield/tpf/lft.py

@@ -94,9 +94,9 @@ def premium(lft_rate: float, selic_over: float) -> float:
         >>> lft_rate = 0.1695 / 100  # 0.1695%
         >>> selic_over = 10.40 / 100  # 10.40%
         >>> yd.lft.premium(lft_rate, selic_over)
-        1.017121
+        1.017120519283759
     """
     adjusted_lft_rate = (lft_rate + 1) * (selic_over + 1) - 1
     f1 = (adjusted_lft_rate + 1) ** (1 / 252) - 1
     f2 = (selic_over + 1) ** (1 / 252) - 1
-    return round(f1 / f2, 6)
+    return f1 / f2

{pyield-0.30.0 → pyield-0.30.2}/pyield/tpf/ltn.py

@@ -175,14 +175,14 @@ def premium(ltn_rate: float, di_rate: float) -> float:
         LTN rate for 01-01-2030: 0.118746
         DI (JAN30) Settlement rate: 0.11725
         >>> yd.ltn.premium(0.118746, 0.11725)
-        1.012072
+        1.0120718007994287
     """
     # Cálculo das taxas diárias
     ltn_factor = (1 + ltn_rate) ** (1 / 252)
     di_factor = (1 + di_rate) ** (1 / 252)
 
     # Retorno do cálculo do prêmio
-    return round((ltn_factor - 1) / (di_factor - 1), 6)
+    return (ltn_factor - 1) / (di_factor - 1)
 
 
 def historical_premium(
@@ -202,7 +202,7 @@ def historical_premium(
 
     Examples:
         >>> yd.ltn.historical_premium("22-08-2024", "01-01-2030")
-        1.012072
+        1.0120718007994287
 
     """
     # Convert input dates to a consistent format

{pyield-0.30.0 → pyield-0.30.2}/pyield/tpf/ntnf.py

@@ -532,7 +532,7 @@ def premium(
         ...     di_expirations=exp_dates,
         ...     di_rates=di_rates,
         ... )
-        1.00996
+        1.0099602136954626
 
     Notes:
         - The function adjusts coupon payment dates to business days and calculates
@@ -572,7 +572,7 @@ def premium(
     factor_ntnf = (1 + ntnf_rate) ** (1 / 252)
     factor_di = (1 + di_ytm) ** (1 / 252)
     premium_np = (factor_ntnf - 1) / (factor_di - 1)
-    return round(float(premium_np), 6)
+    return float(premium_np)
 
 
 def historical_premium(

{pyield-0.30.0 → pyield-0.30.2}/pyproject.toml

@@ -14,7 +14,7 @@ classifiers = [
     "Topic :: Office/Business :: Financial :: Investment",
 ]
 dependencies = [
-    "pandas[html,parquet,excel]>=2.0.0",
+    "pandas[html,parquet]>=2.0.0",
     "numpy>=2.0.0",    
     "requests>=2.31.0",
     "tenacity>=9.0.0",

pyield-0.30.0/pyield/__about__.py

@@ -1 +0,0 @@
-__version__ = "0.30.0"

pyield-0.30.0/pyield/anbima/ipca.py

@@ -1,52 +0,0 @@
-import locale
-from dataclasses import dataclass
-
-import pandas as pd
-
-
-@dataclass
-class IndicatorProjection:
-    reference_period: pd.Period  # Reference month as a pd.Period object
-    projected_value: float  # Projected value
-    last_updated: pd.Timestamp  # Date and time of the last update
-
-
-def ipca_projection() -> IndicatorProjection:
-    """
-    This function retrieves and parses the Excel file that contains economic indicators,
-    specifically looking for the IPCA projection. It extracts the date of the last
-    update and the IPCA projection for the reference month.
-
-    Data file format example after parsing:
-        - ['Data e Hora da Última Atualização: 19/04/2024 - 18:55 h', '', '']
-        - ...
-        - ['IPCA1', 'Projeção (abr/24)', 0.35]
-        - ...
-    """
-    # Define the URL and get the data
-    url = "https://www.anbima.com.br/informacoes/indicadores/arqs/indicadores.xls"
-    df = pd.read_excel(url, dtype_backend="numpy_nullable")
-
-    last_update_str = df.columns[0].split("Atualização:")[-1].strip()
-    last_update = pd.to_datetime(last_update_str, format="%d/%m/%Y - %H:%M h")
-
-    ipca_row = df[df.iloc[:, 0] == "IPCA1"]
-
-    ipca_value = ipca_row.iloc[0, 2]
-    ipca_value = pd.to_numeric(ipca_value, errors="coerce")
-    ipca_value = round(float(ipca_value) / 100, 4)
-
-    # Extract and format the reference month
-    ipca_date = ipca_row.iloc[0, 1]
-    ipca_date = str(ipca_date)
-    ipca_date = ipca_date.split("(")[-1].split(")")[0]
-    locale.setlocale(locale.LC_TIME, "pt_BR.UTF-8")
-    ipca_date = pd.to_datetime(ipca_date, format="%b/%y")
-    reference_period = ipca_date.to_period("M")
-    locale.setlocale(locale.LC_TIME, "")  # Reset locale to default
-
-    return IndicatorProjection(
-        last_updated=last_update,
-        reference_period=reference_period,
-        projected_value=ipca_value,
-    )