修改转换规则 Nanquantile.

paconvert/api_mapping.json

@@ -2087,7 +2087,7 @@
     ]
   },
   "torch.Tensor.nanquantile": {
-    "Matcher": "GenericMatcher",
+    "Matcher": "NanquantileMatcher",
     "paddle_api": "paddle.Tensor.nanquantile",
     "min_input_args": 1,
     "args_list": [

paconvert/api_matcher.py

@@ -4078,3 +4078,20 @@ def generate_code(self, kwargs):
         return ast.parse(
             "paddle.utils.cpp_extension.setup({})".format(self.kwargs_to_str(kwargs))
         )
+
+
+class NanquantileMatcher(BaseMatcher):
+    def generate_code(self, kwargs):
+        new_kwargs = {}
+        kwargs_change = self.api_mapping["kwargs_change"]
+        for k in list(kwargs.keys()):
+            if k in kwargs_change:
+                new_kwargs[kwargs_change[k]] = kwargs[k]
+            else:
+                new_kwargs[k] = kwargs[k]
+            if "q" in k:
+                if "tensor" in kwargs[k] and "[" in kwargs[k]:
+                    new_kwargs[k] = "{}.tolist()".format(kwargs[k])
+
+        code = "{}({})".format(self.get_paddle_api(), self.kwargs_to_str(new_kwargs))
+        return code

tests/test_Tensor_nanquantile.py

@@ -100,3 +100,14 @@ def test_case_7():
         unsupport=True,
         reason="Paddle not support this parameter",
     )
+
+
+def test_multi_q():
+    pytorch_code = textwrap.dedent(
+        """
+        import torch
+        x = torch.tensor([[float('nan'), 1.02, 2.21, 3.333,30, float('nan')]], dtype=torch.float64)
+        result = x.nanquantile(q=torch.tensor([0.3, 0.4], dtype=torch.float64), dim=1, keepdim=True)
+        """
+    )
+    obj.run(pytorch_code, ["result"])