CLN: remove unused in pd._testing

pandas/_testing.py

@@ -32,7 +32,6 @@
     is_datetime64tz_dtype,
     is_extension_array_dtype,
     is_interval_dtype,
-    is_list_like,
     is_number,
     is_period_dtype,
     is_sequence,
@@ -417,10 +416,7 @@ def rands_array(nchars, size, dtype="O"):
         .view((np.str_, nchars))
         .reshape(size)
     )
-    if dtype is None:
-        return retval
-    else:
-        return retval.astype(dtype)
+    return retval.astype(dtype)
 
 
 def randu_array(nchars, size, dtype="O"):
@@ -432,10 +428,7 @@ def randu_array(nchars, size, dtype="O"):
         .view((np.unicode_, nchars))
         .reshape(size)
     )
-    if dtype is None:
-        return retval
-    else:
-        return retval.astype(dtype)
+    return retval.astype(dtype)
 
 
 def rands(nchars):
@@ -448,16 +441,6 @@ def rands(nchars):
     return "".join(np.random.choice(RANDS_CHARS, nchars))
 
 
-def randu(nchars):
-    """
-    Generate one random unicode string.
-
-    See `randu_array` if you want to create an array of random unicode strings.
-
-    """
-    return "".join(np.random.choice(RANDU_CHARS, nchars))
-
-
 def close(fignum=None):
     from matplotlib.pyplot import get_fignums, close as _close
 
@@ -724,10 +707,7 @@ def repr_class(x):
             # return Index as it is to include values in the error message
             return x
 
-        try:
-            return type(x).__name__
-        except AttributeError:
-            return repr(type(x))
+        return type(x).__name__
 
     if exact == "equiv":
         if type(left) != type(right):
@@ -2103,53 +2083,6 @@ def _gen_unique_rand(rng, _extra_size):
     return i.tolist(), j.tolist()
 
 
-def makeMissingCustomDataframe(
-    nrows,
-    ncols,
-    density=0.9,
-    random_state=None,
-    c_idx_names=True,
-    r_idx_names=True,
-    c_idx_nlevels=1,
-    r_idx_nlevels=1,
-    data_gen_f=None,
-    c_ndupe_l=None,
-    r_ndupe_l=None,
-    dtype=None,
-    c_idx_type=None,
-    r_idx_type=None,
-):
-    """
-    Parameters
-    ----------
-    Density : float, optional
-        Float in (0, 1) that gives the percentage of non-missing numbers in
-        the DataFrame.
-    random_state : {np.random.RandomState, int}, optional
-        Random number generator or random seed.
-
-    See makeCustomDataframe for descriptions of the rest of the parameters.
-    """
-    df = makeCustomDataframe(
-        nrows,
-        ncols,
-        c_idx_names=c_idx_names,
-        r_idx_names=r_idx_names,
-        c_idx_nlevels=c_idx_nlevels,
-        r_idx_nlevels=r_idx_nlevels,
-        data_gen_f=data_gen_f,
-        c_ndupe_l=c_ndupe_l,
-        r_ndupe_l=r_ndupe_l,
-        dtype=dtype,
-        c_idx_type=c_idx_type,
-        r_idx_type=r_idx_type,
-    )
-
-    i, j = _create_missing_idx(nrows, ncols, density, random_state)
-    df.values[i, j] = np.nan
-    return df
-
-
 def makeMissingDataframe(density=0.9, random_state=None):
     df = makeDataFrame()
     i, j = _create_missing_idx(*df.shape, density=density, random_state=random_state)
@@ -2397,7 +2330,6 @@ def wrapper(*args, **kwargs):
 def assert_produces_warning(
     expected_warning=Warning,
     filter_level="always",
-    clear=None,
     check_stacklevel=True,
     raise_on_extra_warnings=True,
 ):
@@ -2427,12 +2359,6 @@ class for all warnings. To check that no warning is returned,
           from each module
         * "once" - print the warning the first time it is generated
 
-    clear : str, default None
-        If not ``None`` then remove any previously raised warnings from
-        the ``__warningsregistry__`` to ensure that no warning messages are
-        suppressed by this context manager. If ``None`` is specified,
-        the ``__warningsregistry__`` keeps track of which warnings have been
-        shown, and does not show them again.
     check_stacklevel : bool, default True
         If True, displays the line that called the function containing
         the warning to show were the function is called. Otherwise, the
@@ -2465,19 +2391,6 @@ class for all warnings. To check that no warning is returned,
 
     with warnings.catch_warnings(record=True) as w:
 
-        if clear is not None:
-            # make sure that we are clearing these warnings
-            # if they have happened before
-            # to guarantee that we will catch them
-            if not is_list_like(clear):
-                clear = [clear]
-            for m in clear:
-                try:
-                    m.__warningregistry__.clear()
-                except AttributeError:
-                    # module may not have __warningregistry__
-                    pass
-
         saw_warning = False
         warnings.simplefilter(filter_level)
         yield w