PYield 0.37.4__tar.gz → 0.37.6__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (53) hide show
  1. {pyield-0.37.4 → pyield-0.37.6}/PKG-INFO +1 -1
  2. pyield-0.37.6/pyield/__about__.py +1 -0
  3. {pyield-0.37.4 → pyield-0.37.6}/pyield/anbima/ettj.py +109 -47
  4. {pyield-0.37.4 → pyield-0.37.6}/pyield/anbima/tpf.py +56 -38
  5. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/di1.py +8 -5
  6. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/benchmark.py +78 -59
  7. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/lft.py +6 -15
  8. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/ltn.py +6 -14
  9. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/ntnb.py +6 -15
  10. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/ntnc.py +4 -3
  11. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/ntnf.py +6 -7
  12. pyield-0.37.4/pyield/__about__.py +0 -1
  13. {pyield-0.37.4 → pyield-0.37.6}/.gitignore +0 -0
  14. {pyield-0.37.4 → pyield-0.37.6}/LICENSE +0 -0
  15. {pyield-0.37.4 → pyield-0.37.6}/README.md +0 -0
  16. {pyield-0.37.4 → pyield-0.37.6}/pyield/__init__.py +0 -0
  17. {pyield-0.37.4 → pyield-0.37.6}/pyield/anbima/__init__.py +0 -0
  18. {pyield-0.37.4 → pyield-0.37.6}/pyield/anbima/ima.py +0 -0
  19. {pyield-0.37.4 → pyield-0.37.6}/pyield/anbima/imaq.py +0 -0
  20. {pyield-0.37.4 → pyield-0.37.6}/pyield/anbima/ipca.py +0 -0
  21. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/__init__.py +0 -0
  22. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/di_over.py +0 -0
  23. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/futures/__init__.py +0 -0
  24. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/futures/common.py +0 -0
  25. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/futures/historical.py +0 -0
  26. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/futures/intraday.py +0 -0
  27. {pyield-0.37.4 → pyield-0.37.6}/pyield/b3/futures/xml.py +0 -0
  28. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/__init__.py +0 -0
  29. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/auction.py +0 -0
  30. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/ptax_api.py +0 -0
  31. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/rates.py +0 -0
  32. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/repo.py +0 -0
  33. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/trades_intraday.py +0 -0
  34. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/trades_monthly.py +0 -0
  35. {pyield-0.37.4 → pyield-0.37.6}/pyield/bc/vna.py +0 -0
  36. {pyield-0.37.4 → pyield-0.37.6}/pyield/bday.py +0 -0
  37. {pyield-0.37.4 → pyield-0.37.6}/pyield/conftest.py +0 -0
  38. {pyield-0.37.4 → pyield-0.37.6}/pyield/data_cache.py +0 -0
  39. {pyield-0.37.4 → pyield-0.37.6}/pyield/date_converter.py +0 -0
  40. {pyield-0.37.4 → pyield-0.37.6}/pyield/fwd.py +0 -0
  41. {pyield-0.37.4 → pyield-0.37.6}/pyield/holidays/__init__.py +0 -0
  42. {pyield-0.37.4 → pyield-0.37.6}/pyield/holidays/br_holidays_new.txt +0 -0
  43. {pyield-0.37.4 → pyield-0.37.6}/pyield/holidays/br_holidays_old.txt +0 -0
  44. {pyield-0.37.4 → pyield-0.37.6}/pyield/holidays/brholidays.py +0 -0
  45. {pyield-0.37.4 → pyield-0.37.6}/pyield/ibge/__init__.py +0 -0
  46. {pyield-0.37.4 → pyield-0.37.6}/pyield/ibge/ipca.py +0 -0
  47. {pyield-0.37.4 → pyield-0.37.6}/pyield/interpolator.py +0 -0
  48. {pyield-0.37.4 → pyield-0.37.6}/pyield/py.typed +0 -0
  49. {pyield-0.37.4 → pyield-0.37.6}/pyield/retry.py +0 -0
  50. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/__init__.py +0 -0
  51. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/pre.py +0 -0
  52. {pyield-0.37.4 → pyield-0.37.6}/pyield/tn/tools.py +0 -0
  53. {pyield-0.37.4 → pyield-0.37.6}/pyproject.toml +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: PYield
3
- Version: 0.37.4
3
+ Version: 0.37.6
4
4
  Summary: A Python library for analysis of fixed income instruments in Brazil
5
5
  Project-URL: Source, https://github.com/crdcj/PYield
6
6
  Author-email: Carlos Carvalho <cr.cj@outlook.com>
@@ -0,0 +1 @@
1
+ __version__ = "0.37.6"
@@ -18,7 +18,7 @@ ROUND_DIGITS = 6
18
18
 
19
19
 
20
20
  @default_retry
21
- def _get_last_content() -> str:
21
+ def _get_last_content_text() -> str:
22
22
  """Fetches the raw yield curve data from ANBIMA."""
23
23
  request_payload = {
24
24
  "Idioma": "PT",
@@ -30,34 +30,47 @@ def _get_last_content() -> str:
30
30
  return r.text
31
31
 
32
32
 
33
- def _convert_text_to_df(text: str) -> pd.DataFrame:
34
- """Converts the raw yield curve text to a DataFrame."""
35
- raw_df = pd.read_csv(
36
- StringIO(text),
37
- sep=";",
38
- encoding="latin1",
39
- skiprows=5,
40
- dtype_backend="numpy_nullable",
41
- )
33
+ def _get_reference_date(text: str) -> pd.Timestamp:
34
+ file_date_str = text[0:10] # formato = 09/09/2024
35
+ file_date = pd.to_datetime(file_date_str, format="%d/%m/%Y")
36
+ return file_date
37
+
38
+
39
+ def _filter_ettf_text(texto_completo: str) -> str:
40
+ # Definir os marcadores de início e fim
41
+ marcador_inicio = "Vertices;ETTJ IPCA;ETTJ PREF;Inflação Implícita"
42
+ marcador_fim = "PREFIXADOS (CIRCULAR 3.361)"
43
+
44
+ # 2. Dividir o texto em uma lista de linhas
45
+ linhas = texto_completo.strip().splitlines()
46
+
47
+ # 3. Encontrar os índices das linhas de início e fim
48
+ indice_inicio = linhas.index(marcador_inicio)
49
+ indice_fim = linhas.index(marcador_fim)
50
+
51
+ # 4. Fatiar a lista para extrair o trecho desejado
52
+ trecho_filtrado = linhas[indice_inicio:indice_fim]
53
+
54
+ # Remover linhas vazias que possam ter sido incluídas no final
55
+ while trecho_filtrado and not trecho_filtrado[-1].strip():
56
+ trecho_filtrado.pop()
42
57
 
43
- idx = raw_df.query("Vertices =='PREFIXADOS (CIRCULAR 3.361)'").first_valid_index() # noqa
44
- raw_df.query("index < @idx", inplace=True)
45
- csv_buffer = StringIO()
58
+ # 5. Juntar as linhas filtradas em um único texto e retornar
59
+ return "\n".join(trecho_filtrado)
46
60
 
47
- # Save the DataFrame as CSV in memory buffer to be read again
48
- # It's easier than handling values with commas and dots directly in the DataFrame
49
- raw_df.to_csv(csv_buffer, index=False)
50
- csv_buffer.seek(0) # Move the buffer cursor to the beginning
51
61
 
62
+ def _convert_text_to_df(text: str, reference_date: pd.Timestamp) -> pd.DataFrame:
63
+ """Converts the raw yield curve text to a DataFrame."""
52
64
  df = pd.read_csv(
53
- csv_buffer,
65
+ StringIO(text),
66
+ sep=";",
54
67
  decimal=",",
55
68
  thousands=".",
56
- dtype_backend="numpy_nullable",
69
+ encoding="latin1",
70
+ dtype_backend="pyarrow",
57
71
  )
58
- file_date_str = text[0:10] # 09/09/2024
59
- file_date = pd.to_datetime(file_date_str, format="%d/%m/%Y")
60
- df["date"] = file_date
72
+ df["date"] = reference_date
73
+ df["date"] = df["date"].astype("date32[pyarrow]")
61
74
 
62
75
  return df
63
76
 
@@ -85,9 +98,7 @@ def _process_df(df: pd.DataFrame) -> pd.DataFrame:
85
98
  "real_rate",
86
99
  "implied_inflation",
87
100
  ]
88
- df = df[column_order].copy()
89
-
90
- return df
101
+ return df[column_order].copy()
91
102
 
92
103
 
93
104
  def last_ettj() -> pd.DataFrame:
@@ -111,11 +122,68 @@ def last_ettj() -> pd.DataFrame:
111
122
  Note:
112
123
  All rates are expressed in decimal format (e.g., 0.12 for 12%).
113
124
  """
114
- text = _get_last_content()
115
- df = _convert_text_to_df(text)
125
+ text = _get_last_content_text()
126
+ reference_date = _get_reference_date(text)
127
+ text = _filter_ettf_text(text)
128
+ df = _convert_text_to_df(text, reference_date)
116
129
  return _process_df(df)
117
130
 
118
131
 
132
+ def _extrair_data_e_tabelas(
133
+ texto: str, titulo_tabela: str
134
+ ) -> tuple[pd.Timestamp, str, str]:
135
+ # Títulos que servem como nossos marcadores
136
+ titulo_pre = "ETTJ PREFIXADOS (%a.a./252)"
137
+ titulo_ipca = "ETTJ IPCA (%a.a./252)"
138
+
139
+ # Dividir o texto em linhas
140
+ linhas = [linha for linha in texto.strip().splitlines() if linha.strip()]
141
+
142
+ # Encontrar os índices dos títulos
143
+ inicio_tabela_pre = linhas.index(titulo_pre)
144
+ inicio_tabela_ipca = linhas.index(titulo_ipca)
145
+
146
+ # --- Extrair Tabela 1 ---
147
+ data_ref_pre = linhas[inicio_tabela_pre + 1]
148
+ # A tabela 1 vai do seu cabeçalho até a linha antes do título da tabela 2
149
+ tabela_pre_linhas = linhas[inicio_tabela_pre + 2 : inicio_tabela_ipca]
150
+ tabela_pre = "\n".join(tabela_pre_linhas)
151
+
152
+ # --- Extrair Tabela 2 ---
153
+ data_ref_ipca = linhas[inicio_tabela_ipca + 1]
154
+ # A tabela 2 vai do seu cabeçalho até o final da lista
155
+ tabela_ipca_linhas = linhas[inicio_tabela_ipca + 2 :]
156
+ tabela_ipca = "\n".join(tabela_ipca_linhas)
157
+
158
+ if data_ref_pre != data_ref_ipca:
159
+ raise ValueError("Datas de referência diferentes")
160
+
161
+ data_ref = pd.to_datetime(data_ref_pre, format="%d/%m/%Y")
162
+
163
+ return data_ref, tabela_pre, tabela_ipca
164
+
165
+
166
+ def _ler_tabela_intradia(texto: str) -> pd.DataFrame:
167
+ """
168
+ Reads a table from the given text and returns it as a DataFrame.
169
+
170
+ Args:
171
+ texto (str): The text containing the table data.
172
+
173
+ Returns:
174
+ pd.DataFrame: A DataFrame containing the table data.
175
+ """
176
+ df = pd.read_csv(
177
+ StringIO(texto),
178
+ sep=";",
179
+ decimal=",",
180
+ thousands=".",
181
+ dtype_backend="pyarrow",
182
+ )
183
+ df = df.drop(columns=["Fechamento D -1"])
184
+ return df
185
+
186
+
119
187
  def intraday_ettj() -> pd.DataFrame:
120
188
  """
121
189
  Retrieves and processes the intraday Brazilian yield curve data from ANBIMA.
@@ -140,25 +208,18 @@ def intraday_ettj() -> pd.DataFrame:
140
208
 
141
209
  request_payload = {"Dt_Ref": "", "saida": "csv"}
142
210
  response = requests.post(INTRADAY_ETTJ_URL, data=request_payload)
143
- text_parts = response.text.split("ETTJ IPCA (%a.a./252)")
144
-
145
- nominal_text = text_parts[0]
146
- lines = nominal_text.splitlines()
147
- date_str = lines[1]
148
- date = pd.to_datetime(date_str, format="%d/%m/%Y")
149
- df_text = "\n".join(lines[2:])
150
- df_nominal = pd.read_csv(StringIO(df_text), sep=";", decimal=",", thousands=".")
151
- df_nominal = df_nominal.drop(columns=["Fechamento D -1"])
152
- df_nominal = df_nominal.rename(columns={"D0": "nominal_rate"})
153
-
154
- real_text = text_parts[1]
155
- lines = real_text.splitlines()
156
- df_text = "\n".join(lines[2:])
157
- df_real = pd.read_csv(StringIO(df_text), sep=";", decimal=",", thousands=".")
158
- df_real = df_real.drop(columns=["Fechamento D -1"])
159
- df_real = df_real.rename(columns={"D0": "real_rate"})
160
-
161
- df = pd.merge(df_nominal, df_real, on="Vertices", how="right")
211
+ texto = response.text
212
+
213
+ # --- Extração da Tabela 1: PREFIXADOS ---
214
+ data_ref, tabela_pre, tabela_ipca = _extrair_data_e_tabelas(texto, texto)
215
+
216
+ df_pre = _ler_tabela_intradia(tabela_pre)
217
+ df_pre = df_pre.rename(columns={"D0": "nominal_rate"})
218
+
219
+ df_ipca = _ler_tabela_intradia(tabela_ipca)
220
+ df_ipca = df_ipca.rename(columns={"D0": "real_rate"})
221
+
222
+ df = pd.merge(df_pre, df_ipca, on="Vertices", how="right")
162
223
  df = df.rename(columns={"Vertices": "vertex"})
163
224
 
164
225
  # Divide float columns by 100 and round to 6 decimal places
@@ -167,6 +228,7 @@ def intraday_ettj() -> pd.DataFrame:
167
228
  df["implied_inflation"] = (df["nominal_rate"] + 1) / (df["real_rate"] + 1) - 1
168
229
  df["implied_inflation"] = df["implied_inflation"].round(ROUND_DIGITS)
169
230
 
170
- df["date"] = date
231
+ df["date"] = data_ref
232
+ df["date"] = df["date"].astype("date32[pyarrow]")
171
233
  column_order = ["date", "vertex", "nominal_rate", "real_rate", "implied_inflation"]
172
234
  return df[column_order].copy()
@@ -209,18 +209,19 @@ def _reorder_columns(df: pd.DataFrame) -> pd.DataFrame:
209
209
 
210
210
 
211
211
  def _fetch_tpf_data(date: pd.Timestamp) -> pd.DataFrame:
212
- """Fetch and process TPF secondary market data directly from the ANBIMA source.
212
+ """Busca e processa dados do mercado secundário de TPF diretamente da fonte ANBIMA.
213
213
 
214
- This is a low-level function intended for internal use. It handles the logic
215
- of building the correct URL (public or RTM), downloading the data with retries,
216
- and processing it into a structured DataFrame.
214
+ Esta é uma função de baixo nível para uso interno. Ela lida com a lógica
215
+ de construir a URL correta (pública ou RTM), baixar os dados com novas
216
+ tentativas e processá-los em um DataFrame estruturado.
217
217
 
218
218
  Args:
219
- date (pd.Timestamp): The reference date for the data.
219
+ date (pd.Timestamp): A data de referência para os dados.
220
220
 
221
221
  Returns:
222
- pd.DataFrame: A DataFrame containing processed bond market data, or an
223
- empty DataFrame if data is unavailable or a connection error occurs.
222
+ pd.DataFrame: Um DataFrame contendo os dados de mercado de títulos
223
+ processados, ou um DataFrame vazio se os dados não estiverem
224
+ disponíveis ou ocorrer um erro de conexão.
224
225
  """
225
226
  file_url = _build_file_url(date)
226
227
  date_str = date.strftime("%d/%m/%Y")
@@ -283,38 +284,37 @@ def tpf_data(
283
284
  bond_type: str | None = None,
284
285
  fetch_from_source: bool = False,
285
286
  ) -> pd.DataFrame:
286
- """Retrieve TPF secondary market data from ANBIMA.
287
+ """Recupera os dados do mercado secundário de TPF da ANBIMA.
287
288
 
288
- This function fetches indicative rates and other data for Brazilian government
289
- bonds from ANBIMA. By default, it attempts to retrieve data from a local
290
- cache for performance. If `fetch_from_source` is True, it tries to fetch the data
291
- directly from the ANBIMA website.
289
+ Esta função busca taxas indicativas e outros dados de títulos públicos
290
+ brasileiros. A obtenção dos dados segue uma hierarquia de fontes para
291
+ otimizar o desempenho e o acesso.
292
292
 
293
293
  Args:
294
- date (DateScalar): The reference date for the data (e.g., '2024-06-14').
295
- bond_type (str, optional): Filters results by a specific bond type
296
- (e.g., 'LTN', 'NTN-B'). Defaults to None, returning all bond types.
297
- fetch_from_source (bool, optional): If True, forces the function to
298
- bypass the cache and fetch data directly from the source.
299
- Defaults to False.
294
+ date (DateScalar): A data de referência para os dados (ex: '2024-06-14').
295
+ bond_type (str, optional): Filtra os resultados por um tipo de título
296
+ específico (ex: 'LTN', 'NTN-B'). Por padrão, retorna todos os tipos.
297
+ fetch_from_source (bool, optional): Se True, força a função a ignorar o
298
+ cache e buscar os dados diretamente da fonte (ANBIMA).
299
+ Padrão é False.
300
300
 
301
301
  Returns:
302
- pd.DataFrame: A DataFrame containing the requested bond data. An empty
303
- DataFrame is returned if no data is available for the specified date
304
- (e.g., weekends, holidays, or future dates).
302
+ pd.DataFrame: Um DataFrame contendo os dados solicitados. Retorna um
303
+ DataFrame vazio se não houver dados para a data especificada (ex:
304
+ finais de semana, feriados ou datas futuras).
305
305
 
306
306
  Examples:
307
307
  >>> from pyield import anbima
308
308
  >>> anbima.tpf_data(date="22-08-2025")
309
- ReferenceDate BondType MaturityDate IndicativeRate Price
310
- 0 2025-08-22 LFT 2025-09-01 0.000165 17200.957952
311
- 1 2025-08-22 LFT 2026-03-01 -0.000116 17202.058818
312
- 2 2025-08-22 LFT 2026-09-01 -0.000107 17202.901668
313
- 3 2025-08-22 LFT 2027-03-01 0.000302 17193.200289
314
- 4 2025-08-22 LFT 2027-09-01 0.000411 17186.767105
309
+ ReferenceDate BondType SelicCode ... AskRate IndicativeRate DIRate
310
+ 0 2025-08-22 LFT 210100 ... 0.0001 0.000165 0.14906
311
+ 1 2025-08-22 LFT 210100 ... -0.000156 -0.000116 0.14843
312
+ 2 2025-08-22 LFT 210100 ... -0.000143 -0.000107 0.1436
313
+ 3 2025-08-22 LFT 210100 ... 0.000292 0.000302 0.138189
315
314
  ...
316
315
 
317
- DataFrame columns:
316
+
317
+ Data columns:
318
318
  - BondType: Tipo do título público (e.g., 'LTN', 'NTN-B').
319
319
  - ReferenceDate: Data de referência dos dados.
320
320
  - SelicCode: Código do título no SELIC.
@@ -329,7 +329,7 @@ def tpf_data(
329
329
  - BidRate: Taxa de compra em formato decimal (e.g., 0.10 para 10%).
330
330
  - AskRate: Taxa de venda em formato decimal.
331
331
  - IndicativeRate: Taxa indicativa em formato decimal.
332
- - DIRate: Taxa DI interpolada para o vencimento do título.
332
+ - DIRate: Taxa DI interpolada (flatforward) no vencimento do título.
333
333
  - StdDev: Desvio padrão da taxa indicativa.
334
334
  - LowerBoundRateD0: Limite inferior do intervalo indicativo para D+0.
335
335
  - UpperBoundRateD0: Limite superior do intervalo indicativo para D+0.
@@ -338,11 +338,16 @@ def tpf_data(
338
338
  - Criteria: Critério utilizado pela ANBIMA para o cálculo.
339
339
 
340
340
  Notes:
341
- - Data for the last 5 business days is available on the public ANBIMA
342
- website.
343
- - Historical data (older than 5 business days) is only available via
344
- the RTM network. If you are not connected to RTM, requests for
345
- older dates will return an empty DataFrame.
341
+ A fonte dos dados segue a seguinte hierarquia:
342
+
343
+ 1. **Cache Local (Padrão):** Fornece acesso rápido a dados históricos
344
+ desde 01/01/2020. É utilizado por padrão (`fetch_from_source=False`).
345
+ 2. **Site Público da ANBIMA:** Acessado quando `fetch_from_source=True`,
346
+ disponibiliza os dados dos últimos 5 dias úteis.
347
+ 3. **Rede RTM da ANBIMA:** Acessada quando `fetch_from_source=True` para
348
+ datas com mais de 5 dias úteis. O acesso ao histórico completo
349
+ requer uma conexão à rede RTM. Sem ela, a consulta para datas
350
+ antigas retornará um DataFrame vazio.
346
351
  """
347
352
  date = convert_input_dates(date)
348
353
 
@@ -377,11 +382,10 @@ def tpf_data(
377
382
 
378
383
 
379
384
  def tpf_fixed_rate_maturities(date: DateScalar) -> pd.Series:
380
- """Retrieve pre-defined maturity dates for LTN and NTN-F bonds.
385
+ """Retrieve existing maturity dates for fixed-rate public bonds (LTN and NTN-F).
381
386
 
382
- This function fetches pre-defined maturity dates for 'LTN' (prefixadas) and
383
- 'NTN-F' (indexadas ao CDI) bond types from the cached ANBIMA dataset
384
- for a given reference date.
387
+ This function fetches existing maturity dates for 'LTN' and
388
+ 'NTN-F' bond types for a given date.
385
389
 
386
390
  Args:
387
391
  date (DateScalar): The reference date for maturity dates.
@@ -389,6 +393,20 @@ def tpf_fixed_rate_maturities(date: DateScalar) -> pd.Series:
389
393
  Returns:
390
394
  pd.Series: A Series containing unique maturity dates for 'LTN' and
391
395
  'NTN-F' bonds, sorted in ascending order.
396
+
397
+ Examples:
398
+ >>> from pyield import anbima
399
+ >>> anbima.tpf_fixed_rate_maturities(date="22-08-2025")
400
+ 0 2025-10-01
401
+ 1 2026-01-01
402
+ 2 2026-04-01
403
+ 3 2026-07-01
404
+ ...
405
+ 14 2031-01-01
406
+ 15 2032-01-01
407
+ 16 2033-01-01
408
+ 17 2035-01-01
409
+ Name: MaturityDate, dtype: date32[day][pyarrow]
392
410
  """
393
411
  maturity_dates = (
394
412
  tpf_data(date)
@@ -37,11 +37,14 @@ def _get_data(dates: pd.Series) -> pd.DataFrame:
37
37
  ):
38
38
  try:
39
39
  df_intraday = b3.futures(contract_code="DI1", date=today)
40
- if not df_intraday.empty:
41
- logger.info("Dados intraday de DI inseridos com sucesso.")
42
- return pd.concat([df_cached, df_intraday]).reset_index(drop=True)
43
- else:
44
- logger.warning(f"Nenhum dado intraday de DI encontrado para {today}.")
40
+ if "SettlementPrice" not in df_intraday.columns or df_intraday.empty:
41
+ logger.warning(
42
+ f"Ainda sem dados de ajustes intraday para {today}."
43
+ " Retornando apenas dados do cache."
44
+ )
45
+ return df_cached
46
+ logger.info("Dados intraday de DI inseridos com sucesso.")
47
+ return pd.concat([df_cached, df_intraday]).reset_index(drop=True)
45
48
  except Exception as e:
46
49
  logger.error(f"Falha ao buscar dados intraday para {today}: {e}")
47
50
 
@@ -7,8 +7,6 @@ import requests
7
7
 
8
8
  logger = logging.getLogger(__name__)
9
9
 
10
- TIMEZONE_BZ = ZoneInfo("America/Sao_Paulo")
11
-
12
10
  COLUMN_MAPPING = {
13
11
  "INÍCIO": "StartDate",
14
12
  "TERMINO": "EndDate",
@@ -17,14 +15,71 @@ COLUMN_MAPPING = {
17
15
  "BENCHMARK": "Benchmark",
18
16
  }
19
17
 
20
-
21
18
  API_BASE_URL = (
22
19
  "https://apiapex.tesouro.gov.br/aria/v1/api-leiloes-pub/custom/benchmarks"
23
20
  )
24
21
  API_HISTORY_PARAM = "incluir_historico"
25
22
 
23
+ TIMEZONE_BZ = ZoneInfo("America/Sao_Paulo")
24
+
25
+
26
+ def _fetch_raw_benchmarks(include_history: bool) -> list[dict]:
27
+ """
28
+ Fetches the raw benchmark data from the Tesouro Nacional API.
29
+ Handles network requests, retries, and basic response validation.
30
+ """
31
+ session = requests.Session()
32
+ include_history_param_value = "S" if include_history else "N"
33
+ api_endpoint = f"{API_BASE_URL}?{API_HISTORY_PARAM}={include_history_param_value}"
34
+
35
+ try:
36
+ response = session.get(api_endpoint)
37
+ response.raise_for_status()
38
+ except requests.exceptions.SSLError as e:
39
+ logger.warning(
40
+ f"SSL error encountered: {e}. Retrying without certificate verification."
41
+ )
42
+ response = session.get(api_endpoint, verify=False)
43
+ response.raise_for_status()
44
+ except requests.exceptions.RequestException as e:
45
+ logger.error(f"Error fetching benchmarks from API: {e}")
46
+ return [] # Retorna lista vazia em caso de erro
47
+
48
+ response_dict = response.json()
49
+
50
+ if not response_dict or "registros" not in response_dict:
51
+ logger.warning("API response did not contain 'registros' key or was empty.")
52
+ return []
53
+
54
+ return response_dict["registros"]
55
+
56
+
57
+ def _process_benchmark_data(raw_data: list[dict]) -> pd.DataFrame:
58
+ """
59
+ Converts raw benchmark data into a cleaned and typed pandas DataFrame.
60
+ """
61
+ if not raw_data:
62
+ return pd.DataFrame()
63
+
64
+ df = pd.DataFrame(raw_data).dropna().convert_dtypes(dtype_backend="pyarrow")
26
65
 
27
- def benchmarks(include_history: bool = False) -> pd.DataFrame:
66
+ if df.empty:
67
+ logger.warning("DataFrame is empty after dropping NaN values.")
68
+ return pd.DataFrame()
69
+
70
+ # Limpeza e tipagem
71
+ date_cols = ["VENCIMENTO", "TERMINO", "INÍCIO"]
72
+ for col in date_cols:
73
+ df[col] = pd.to_datetime(df[col]).astype("date32[pyarrow]")
74
+
75
+ str_cols = ["BENCHMARK", "TÍTULO"]
76
+ for col in str_cols:
77
+ df[col] = df[col].str.strip()
78
+
79
+ return df
80
+
81
+
82
+ def benchmarks(bond_type: str = None, include_history: bool = False) -> pd.DataFrame:
28
83
  """Fetches benchmark data for Brazilian Treasury Bonds from the TN API.
29
84
 
30
85
  This function retrieves current or historical benchmark data for various Brazilian
@@ -56,9 +111,8 @@ def benchmarks(include_history: bool = False) -> pd.DataFrame:
56
111
  Examples:
57
112
  >>> from pyield import tn
58
113
  >>> df_current = tn.benchmarks()
59
-
60
114
  >>> # Get historical benchmarks
61
- >>> df_history = tn.benchmarks(include_history=True)
115
+ >>> df_history = tn.benchmarks(bond_type="LFT", include_history=True)
62
116
  >>> df_history.head()
63
117
  StartDate EndDate BondType MaturityDate Benchmark
64
118
  0 2014-01-01 2014-06-30 LFT 2020-03-01 LFT 6 anos
@@ -67,67 +121,32 @@ def benchmarks(include_history: bool = False) -> pd.DataFrame:
67
121
  3 2015-05-01 2015-12-31 LFT 2021-09-01 LFT 6 anos
68
122
  4 2016-01-01 2016-06-30 LFT 2022-03-01 LFT 6 anos
69
123
  """
70
- session = requests.Session()
71
- include_history_param_value = "S" if include_history else "N"
72
- api_endpoint = f"{API_BASE_URL}?{API_HISTORY_PARAM}={include_history_param_value}"
73
-
74
- try:
75
- stn_benchmarks = session.get(api_endpoint)
76
- stn_benchmarks.raise_for_status()
77
- except requests.exceptions.SSLError as e:
78
- logger.error(
79
- f"SSL error encountered: {e}. Retrying without certificate verification."
80
- )
81
- stn_benchmarks = session.get(api_endpoint, verify=False)
82
- stn_benchmarks.raise_for_status()
83
- except requests.exceptions.RequestException as e:
84
- logger.error(f"Error fetching benchmarks from API: {e}")
85
- return pd.DataFrame()
124
+ # Passo 1: Buscar os dados brutos
125
+ raw_data = _fetch_raw_benchmarks(include_history=include_history)
126
+ if not raw_data:
127
+ return pd.DataFrame() # Retorna DF vazio se a busca falhou
86
128
 
87
- response_dict = stn_benchmarks.json()
88
-
89
- # 1a verificação: Verifica se a resposta da API contém dados válidos de registros
90
- if not response_dict or not response_dict.get("registros"):
91
- logger.warning("API response did not contain 'registros' key or it was empty.")
92
- return pd.DataFrame()
93
-
94
- # Tenta criar o DataFrame.
95
- df = (
96
- pd.DataFrame(response_dict["registros"])
97
- .dropna()
98
- .convert_dtypes(dtype_backend="pyarrow")
99
- )
100
-
101
- # 2a verificação: Verifica se o DataFrame resultante (pós-dropna) está vazio
102
- # Esta verificação é importante porque .dropna() pode remover todas as linhas.
129
+ # Passo 2: Processar os dados em um DataFrame limpo
130
+ df = _process_benchmark_data(raw_data)
103
131
  if df.empty:
104
- logger.warning(
105
- "No valid benchmark data found after initial processing"
106
- "(e.g., all rows had NaNs and were dropped).",
107
- )
108
- return pd.DataFrame()
109
-
110
- # Se chegamos até aqui, o DataFrame tem dados e pode ser processado.
111
- for col in ["VENCIMENTO", "TERMINO", "INÍCIO"]:
112
- df[col] = pd.to_datetime(df[col]).astype("date32[pyarrow]")
113
-
114
- df["BENCHMARK"] = df["BENCHMARK"].str.strip()
115
- df["TÍTULO"] = df["TÍTULO"].str.strip()
132
+ return pd.DataFrame() # Retorna DF vazio se o processamento resultou em nada
116
133
 
134
+ # Passo 3: Aplicar filtros específicos da chamada
117
135
  if not include_history:
118
- # Em tese, a API já retorna apenas benchmarks ativos,
119
- # mas vamos garantir que o DataFrame só contenha benchmarks ativos
120
- # considerando o período atual.
121
136
  today = dt.now(TIMEZONE_BZ).date() # noqa
122
137
  df = df.query("INÍCIO <= @today <= TERMINO").reset_index(drop=True)
138
+ if df.empty:
139
+ logger.warning(
140
+ "No current benchmark data found after filtering by active period."
141
+ )
123
142
 
124
- # Verifica novamente se o DataFrame ficou vazio *após o filtro condicional*
125
- # (apenas se `include_history` for False)
126
- if df.empty and not include_history:
127
- logger.warning(
128
- "No current benchmark data found after filtering by active period."
129
- )
143
+ if bond_type:
144
+ df = df.query("TÍTULO == @bond_type").reset_index(drop=True)
145
+
146
+ if df.empty:
147
+ return pd.DataFrame() # Retorna DF vazio se os filtros removeram tudo
130
148
 
149
+ # Passo 4: Formatar a saída final
131
150
  column_order = [c for c in COLUMN_MAPPING if c in df.columns]
132
151
  return (
133
152
  df[column_order]
@@ -24,21 +24,12 @@ def data(date: DateScalar) -> pd.DataFrame:
24
24
  Examples:
25
25
  >>> from pyield import lft
26
26
  >>> lft.data("23-08-2024")
27
- ReferenceDate BondType MaturityDate IndicativeRate Price
28
- 0 2024-08-23 LFT 2024-09-01 0.000272 15252.158852
29
- 1 2024-08-23 LFT 2025-03-01 -0.000418 15255.605864
30
- 2 2024-08-23 LFT 2025-09-01 -0.00023 15255.819395
31
- 3 2024-08-23 LFT 2026-03-01 0.000075 15250.526859
32
- 4 2024-08-23 LFT 2026-09-01 0.000114 15248.757596
33
- 5 2024-08-23 LFT 2027-03-01 0.000669 15226.824838
34
- 6 2024-08-23 LFT 2027-09-01 0.000948 15208.842417
35
- 7 2024-08-23 LFT 2028-03-01 0.001172 15189.853347
36
- 8 2024-08-23 LFT 2028-09-01 0.001328 15171.352348
37
- 9 2024-08-23 LFT 2029-03-01 0.001491 15150.700781
38
- 10 2024-08-23 LFT 2029-09-01 0.001587 15131.894737
39
- 11 2024-08-23 LFT 2030-03-01 0.001591 15119.952213
40
- 12 2024-08-23 LFT 2030-06-01 0.001641 15109.717943
41
- 13 2024-08-23 LFT 2030-09-01 0.001687 15099.285393
27
+ ReferenceDate BondType SelicCode ... AskRate IndicativeRate DIRate
28
+ 0 2024-08-23 LFT 210100 ... 0.000226 0.000272 0.10408
29
+ 1 2024-08-23 LFT 210100 ... -0.000481 -0.000418 0.11082
30
+ 2 2024-08-23 LFT 210100 ... -0.000258 -0.00023 0.114315
31
+ 3 2024-08-23 LFT 210100 ... 0.00006 0.000075 0.114982
32
+ ...
42
33
  """
43
34
  return anbima.tpf_data(date, "LFT")
44
35
 
@@ -22,20 +22,12 @@ def data(date: DateScalar) -> pd.DataFrame:
22
22
  Examples:
23
23
  >>> from pyield import ltn
24
24
  >>> ltn.data("23-08-2024")
25
- ReferenceDate BondType MaturityDate IndicativeRate Price
26
- 0 2024-08-23 LTN 2024-10-01 0.104416 989.415342
27
- 1 2024-08-23 LTN 2025-01-01 0.107171 964.293046
28
- 2 2024-08-23 LTN 2025-04-01 0.110866 938.943013
29
- 3 2024-08-23 LTN 2025-07-01 0.113032 913.849158
30
- 4 2024-08-23 LTN 2025-10-01 0.114374 887.394285
31
- 5 2024-08-23 LTN 2026-01-01 0.114654 863.026594
32
- 6 2024-08-23 LTN 2026-04-01 0.114997 840.232741
33
- 7 2024-08-23 LTN 2026-07-01 0.115265 818.020491
34
- 8 2024-08-23 LTN 2026-10-01 0.115357 795.185488
35
- 9 2024-08-23 LTN 2027-07-01 0.115335 733.981131
36
- 10 2024-08-23 LTN 2028-01-01 0.115694 693.647778
37
- 11 2024-08-23 LTN 2028-07-01 0.116417 655.6398
38
- 12 2024-08-23 LTN 2030-01-01 0.117436 554.331151
25
+ ReferenceDate BondType SelicCode ... AskRate IndicativeRate DIRate
26
+ 0 2024-08-23 LTN 100000 ... 0.104252 0.104416 0.10472
27
+ 1 2024-08-23 LTN 100000 ... 0.107016 0.107171 0.10823
28
+ 2 2024-08-23 LTN 100000 ... 0.110746 0.110866 0.11179
29
+ 3 2024-08-23 LTN 100000 ... 0.112947 0.113032 0.11365
30
+ ...
39
31
 
40
32
  """
41
33
  return anbima.tpf_data(date, "LTN")
@@ -36,21 +36,12 @@ def data(date: DateScalar) -> pd.DataFrame:
36
36
  Examples:
37
37
  >>> from pyield import ntnb
38
38
  >>> ntnb.data("23-08-2024")
39
- ReferenceDate BondType MaturityDate IndicativeRate Price
40
- 0 2024-08-23 NTN-B 2025-05-15 0.063804 4377.008543
41
- 1 2024-08-23 NTN-B 2026-08-15 0.065795 4278.316344
42
- 2 2024-08-23 NTN-B 2027-05-15 0.063794 4350.54878
43
- 3 2024-08-23 NTN-B 2028-08-15 0.063094 4281.186307
44
- 4 2024-08-23 NTN-B 2029-05-15 0.0621 4358.101314
45
- 5 2024-08-23 NTN-B 2030-08-15 0.060298 4324.468801
46
- 6 2024-08-23 NTN-B 2032-08-15 0.0604 4320.153067
47
- 7 2024-08-23 NTN-B 2033-05-15 0.060403 4384.189924
48
- 8 2024-08-23 NTN-B 2035-05-15 0.060375 4386.002533
49
- 9 2024-08-23 NTN-B 2040-08-15 0.059797 4345.119587
50
- 10 2024-08-23 NTN-B 2045-05-15 0.060923 4358.235102
51
- 11 2024-08-23 NTN-B 2050-08-15 0.06114 4279.434119
52
- 12 2024-08-23 NTN-B 2055-05-15 0.060892 4355.145036
53
- 13 2024-08-23 NTN-B 2060-08-15 0.061005 4282.308398
39
+ ReferenceDate BondType SelicCode ... AskRate IndicativeRate DIRate
40
+ 0 2024-08-23 NTN-B 760199 ... 0.063667 0.063804 0.112749
41
+ 1 2024-08-23 NTN-B 760199 ... 0.065635 0.065795 0.114963
42
+ 2 2024-08-23 NTN-B 760199 ... 0.063601 0.063794 0.114888
43
+ 3 2024-08-23 NTN-B 760199 ... 0.062905 0.063094 0.115595
44
+ ...
54
45
  """
55
46
  return anbima.tpf_data(date, "NTN-B")
56
47
 
@@ -52,9 +52,10 @@ def data(date: DateScalar) -> pd.DataFrame:
52
52
 
53
53
  Examples:
54
54
  >>> from pyield import ntnc
55
- >>> ntnc.data("21-03-2025")
56
- ReferenceDate BondType MaturityDate IndicativeRate Price
57
- 0 2025-03-21 NTN-C 2031-01-01 0.067626 8347.348705
55
+ >>> ntnc.data("23-08-2024")
56
+ ReferenceDate BondType SelicCode ... AskRate IndicativeRate DIRate
57
+ 0 2024-08-23 NTN-C 770100 ... 0.057587 0.059617 0.11575
58
+ ...
58
59
  """
59
60
  return anbima.tpf_data(date, "NTN-C")
60
61
 
@@ -36,13 +36,12 @@ def data(date: DateScalar) -> pd.DataFrame:
36
36
  Examples:
37
37
  >>> from pyield import ntnf
38
38
  >>> ntnf.data("23-08-2024")
39
- ReferenceDate BondType MaturityDate IndicativeRate Price
40
- 0 2024-08-23 NTN-F 2025-01-01 0.107692 1011.189166
41
- 1 2024-08-23 NTN-F 2027-01-01 0.115109 985.834842
42
- 2 2024-08-23 NTN-F 2029-01-01 0.116337 964.126325
43
- 3 2024-08-23 NTN-F 2031-01-01 0.117008 945.416939
44
- 4 2024-08-23 NTN-F 2033-01-01 0.116307 934.776692
45
- 5 2024-08-23 NTN-F 2035-01-01 0.116586 923.239406
39
+ ReferenceDate BondType SelicCode ... AskRate IndicativeRate DIRate
40
+ 0 2024-08-23 NTN-F 950199 ... 0.107524 0.107692 0.10823
41
+ 1 2024-08-23 NTN-F 950199 ... 0.114948 0.115109 0.11467
42
+ 2 2024-08-23 NTN-F 950199 ... 0.11621 0.116337 0.1156
43
+ 3 2024-08-23 NTN-F 950199 ... 0.116958 0.117008 0.11575
44
+ ...
46
45
  """
47
46
  return anbima.tpf_data(date, "NTN-F")
48
47
 
@@ -1 +0,0 @@
1
- __version__ = "0.37.4"
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes