DOC: Fix documentation for ADF and related tests

arch/tests/unitroot/test_unitroot.py

@@ -432,8 +432,22 @@ def test_representations(trend):
 def test_unknown_method():
     rnd = np.random.RandomState(12345)
     y = np.cumsum(rnd.standard_normal(250))
-    with pytest.raises(ValueError, match=r"Unknown method"):
+    with pytest.raises(ValueError, match=r"method must be one"):
         assert np.isfinite(ADF(y, method="unknown").stat)
+    with pytest.raises(TypeError, match=r"method must be a string"):
+        assert np.isfinite(ADF(y, method=1).stat)
+
+
+@pytest.mark.parametrize("method", ["aic", "AIC", "BIC", "bic", "t-stat", "tstat", "t"])
+def test_acceptable_method(method):
+    rnd = np.random.RandomState(12345)
+    y = np.cumsum(rnd.standard_normal(250))
+    used_method = ADF(y, method=method)._method
+    if method in ("t-stat", "t", "tstat"):
+        expected_method = "t-stat"
+    else:
+        expected_method = method.lower()
+    assert used_method == expected_method
 
 
 def test_auto_low_memory():
@@ -593,12 +607,13 @@ def test_zivot_andrews(series_name):
     # Test results from package urca.ur.za (1.13-0)
     y = ZIVOT_ANDREWS_DATA[series_name].dropna()
     result = series[series_name]
+    kwargs = {"method": result.method} if result.method is not None else {}
     za = ZivotAndrews(
         y,
         lags=result.lags,
         trend=result.trend,
         max_lags=result.max_lags,
-        method=result.method,
+        **kwargs,
     )
     assert_almost_equal(za.stat, result.stat, decimal=3)
     assert_almost_equal(za.pvalue, result.pvalue, decimal=3)

arch/unitroot/unitroot.py

@@ -187,8 +187,6 @@ def _select_best_ic(
         large_tstat = abs(tstat) >= stop
         lag = int(squeeze(argwhere(large_tstat)).max())
         icbest = float(tstat[lag])
-    else:
-        raise ValueError("Unknown method")
 
     return icbest, lag
 
@@ -548,6 +546,16 @@ def _compute_if_needed(self) -> None:
             self._check_specification()
             self._compute_statistic()
 
+    def _clean_method(self, method: str) -> Literal["aic", "bic", "t-stat"]:
+        if not isinstance(method, str):
+            raise TypeError("method must be a string")
+        method = method.lower()
+        if method in ("t", "tstat", "t-stat"):
+            return "t-stat"
+        elif method not in ("aic", "bic"):
+            raise ValueError("method must be one of 'aic', 'bic' or 't-stat'")
+        return method
+
     @property
     def null_hypothesis(self) -> str:
         """The null hypothesis"""
@@ -683,11 +691,11 @@ class ADF(UnitRootTest, metaclass=AbstractDocStringInheritor):
 
     max_lags : int, optional
         The maximum number of lags to use when selecting lag length
-    method : {"AIC", "BIC", "t-stat"}, optional
+    method : {"aic", "bic", "t-stat"}, optional
         The method to use when selecting the lag length
 
-        - "AIC" - Select the minimum of the Akaike IC
-        - "BIC" - Select the minimum of the Schwarz/Bayesian IC
+        - "aic" - Select the minimum of the Akaike IC
+        - "bic" - Select the minimum of the Schwarz/Bayesian IC
         - "t-stat" - Select the minimum of the Schwarz/Bayesian IC
 
     low_memory : bool
@@ -763,7 +771,7 @@ def __init__(
         valid_trends = ("n", "c", "ct", "ctt")
         super().__init__(y, lags, trend, valid_trends)
         self._max_lags = max_lags
-        self._method = method
+        self._method = self._clean_method(method)
         self._test_name = "Augmented Dickey-Fuller"
         self._regression = None
         self._low_memory = bool(low_memory)
@@ -798,7 +806,7 @@ def _compute_statistic(self) -> None:
         y, trend, lags = self._y, self._trend, self._lags
         resols = _estimate_df_regression(y, cast("UnitRootTrend", trend), lags)
         self._regression = resols
-        (self._stat, *_) = (stat, *_) = resols.tvalues
+        self._stat, *_ = (stat, *_) = resols.tvalues
         self._nobs = int(resols.nobs)
         self._pvalue = mackinnonp(
             stat,
@@ -851,11 +859,11 @@ class DFGLS(UnitRootTest, metaclass=AbstractDocStringInheritor):
         The maximum number of lags to use when selecting lag length. When using
         automatic lag length selection, the lag is selected using OLS
         detrending rather than GLS detrending ([pq]_).
-    method : {"AIC", "BIC", "t-stat"}, optional
+    method : {"aic", "bic", "t-stat"}, optional
         The method to use when selecting the lag length
 
-        - "AIC" - Select the minimum of the Akaike IC
-        - "BIC" - Select the minimum of the Schwarz/Bayesian IC
+        - "aic" - Select the minimum of the Akaike IC
+        - "bic" - Select the minimum of the Schwarz/Bayesian IC
         - "t-stat" - Select the minimum of the Schwarz/Bayesian IC
 
     Notes
@@ -912,7 +920,7 @@ def __init__(
         valid_trends = ("c", "ct")
         super().__init__(y, lags, trend, valid_trends)
         self._max_lags = max_lags
-        self._method = method
+        self._method = self._clean_method(method)
         self._regression = None
         self._low_memory = low_memory
         if low_memory is None:
@@ -1403,11 +1411,11 @@ class ZivotAndrews(UnitRootTest, metaclass=AbstractDocStringInheritor):
         calculation in range [0, 0.333] (default=0.15)
     max_lags : int, optional
         The maximum number of lags to use when selecting lag length
-    method : {"AIC", "BIC", "t-stat"}, optional
+    method : {"aic", "bic", "t-stat"}, optional
         The method to use when selecting the lag length
 
-        - "AIC" - Select the minimum of the Akaike IC
-        - "BIC" - Select the minimum of the Schwarz/Bayesian IC
+        - "aic" - Select the minimum of the Akaike IC
+        - "bic" - Select the minimum of the Schwarz/Bayesian IC
         - "t-stat" - Select the minimum of the Schwarz/Bayesian IC
 
     Notes
@@ -1457,7 +1465,8 @@ def __init__(
             raise ValueError("trim must be a float in range [0, 1/3]")
         self._trim = trim
         self._max_lags = max_lags
-        self._method = method
+        self._method = self._clean_method(method)
+
         self._test_name = "Zivot-Andrews"
         self._all_stats = full(self._y.shape[0], nan)
         self._null_hypothesis = (