[FIX] Contiguity on exogenous (#591)

Co-authored-by: José Morales <jmoralz92@gmail.com>

nbs/src/nixtla_client.ipynb

@@ -798,22 +798,29 @@
     "        payload: dict[str, Any],\n",
     "        multithreaded_compress: bool,\n",
     "    ) -> dict[str, Any]:\n",
+    "        def ensure_contiguous_if_array(x):\n",
+    "            if not isinstance(x, np.ndarray):\n",
+    "                return x\n",
+    "            if np.issubdtype(x.dtype, np.floating):\n",
+    "                x = np.nan_to_num(\n",
+    "                    np.ascontiguousarray(x, dtype=np.float32),\n",
+    "                    nan=np.nan,\n",
+    "                    posinf=np.finfo(np.float32).max,\n",
+    "                    neginf=np.finfo(np.float32).min,\n",
+    "                    copy=False,\n",
+    "                )\n",
+    "            else:\n",
+    "                x = np.ascontiguousarray(x)\n",
+    "            return x\n",
+    "\n",
     "        def ensure_contiguous_arrays(d: dict[str, Any]) -> None:\n",
     "            for k, v in d.items():\n",
     "                if isinstance(v, np.ndarray):\n",
-    "                    if np.issubdtype(v.dtype, np.floating):\n",
-    "                        v_cont = np.ascontiguousarray(v, dtype=np.float32)\n",
-    "                        d[k] = np.nan_to_num(\n",
-    "                            v_cont, \n",
-    "                            nan=np.nan, \n",
-    "                            posinf=np.finfo(np.float32).max, \n",
-    "                            neginf=np.finfo(np.float32).min,\n",
-    "                            copy=False,\n",
-    "                        )\n",
-    "                    else:\n",
-    "                        d[k] = np.ascontiguousarray(v)\n",
+    "                    d[k] = ensure_contiguous_if_array(v)\n",
+    "                elif isinstance(v, list):\n",
+    "                    d[k] = [ensure_contiguous_if_array(x) for x in v]                    \n",
     "                elif isinstance(v, dict):\n",
-    "                    ensure_contiguous_arrays(v) \n",
+    "                    ensure_contiguous_arrays(v)\n",
     "\n",
     "        ensure_contiguous_arrays(payload)\n",
     "        content = orjson.dumps(payload, option=orjson.OPT_SERIALIZE_NUMPY)\n",
@@ -3369,6 +3376,27 @@
     "        h=7,\n",
     "        add_history=True,\n",
     "        num_partitions=2,\n",
+    "    )\n",
+    "    df_freq[\"exog_1\"] = 1\n",
+    "    test_num_partitions_same_results(\n",
+    "        nixtla_client.detect_anomalies,\n",
+    "        level=98,\n",
+    "        df=df_freq,\n",
+    "        num_partitions=2,\n",
+    "    )\n",
+    "    test_num_partitions_same_results(\n",
+    "        nixtla_client.cross_validation,\n",
+    "        h=7,\n",
+    "        n_windows=2,\n",
+    "        df=df_freq,\n",
+    "        num_partitions=2,\n",
+    "    )\n",
+    "    test_num_partitions_same_results(\n",
+    "        nixtla_client.forecast,\n",
+    "        df=df_freq,\n",
+    "        h=7,\n",
+    "        add_history=True,\n",
+    "        num_partitions=2,\n",
     "    )"
    ]
   },

nixtla/__init__.py

@@ -1,3 +1,3 @@
-__version__ = "0.6.5"
+__version__ = "0.6.6"
 __all__ = ["NixtlaClient"]
 from .nixtla_client import NixtlaClient

nixtla/nixtla_client.py

@@ -723,20 +723,27 @@ def _make_request(
         payload: dict[str, Any],
         multithreaded_compress: bool,
     ) -> dict[str, Any]:
+        def ensure_contiguous_if_array(x):
+            if not isinstance(x, np.ndarray):
+                return x
+            if np.issubdtype(x.dtype, np.floating):
+                x = np.nan_to_num(
+                    np.ascontiguousarray(x, dtype=np.float32),
+                    nan=np.nan,
+                    posinf=np.finfo(np.float32).max,
+                    neginf=np.finfo(np.float32).min,
+                    copy=False,
+                )
+            else:
+                x = np.ascontiguousarray(x)
+            return x
+
         def ensure_contiguous_arrays(d: dict[str, Any]) -> None:
             for k, v in d.items():
                 if isinstance(v, np.ndarray):
-                    if np.issubdtype(v.dtype, np.floating):
-                        v_cont = np.ascontiguousarray(v, dtype=np.float32)
-                        d[k] = np.nan_to_num(
-                            v_cont,
-                            nan=np.nan,
-                            posinf=np.finfo(np.float32).max,
-                            neginf=np.finfo(np.float32).min,
-                            copy=False,
-                        )
-                    else:
-                        d[k] = np.ascontiguousarray(v)
+                    d[k] = ensure_contiguous_if_array(v)
+                elif isinstance(v, list):
+                    d[k] = [ensure_contiguous_if_array(x) for x in v]
                 elif isinstance(v, dict):
                     ensure_contiguous_arrays(v)
 

settings.ini

@@ -8,7 +8,7 @@ author = Nixtla
 author_email = business@nixtla.io
 copyright = Nixtla Inc.
 branch = main
-version = 0.6.5
+version = 0.6.6
 min_python = 3.9
 audience = Developers
 language = English
@@ -17,7 +17,7 @@ license = apache2
 status = 4
 requirements = annotated-types httpx[zstd] orjson pandas pydantic>=1.10 tenacity tqdm utilsforecast>=0.2.8
 dev_requirements = black datasetsforecast fire hierarchicalforecast ipywidgets jupyterlab nbdev neuralforecast numpy<2 plotly polars pre-commit pyreadr python-dotenv pyyaml setuptools<70 statsforecast tabulate
-distributed_requirements = fugue[dask,ray,spark]>=0.8.7 pandas<2.2 ray<2.6.3
+distributed_requirements = fugue[dask,ray,spark]>=0.8.7 dask<=2024.12.1 pandas<2.2 ray<2.6.3
 plotting_requirements = utilsforecast[plotting]
 date_extra_requirements = holidays
 nbs_path = nbs