data -> value

Author: Matt Sokoloff

examples/annotation_types/basics.ipynb

@@ -33,7 +33,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install \"labelbox[data]==3.0.0rc1\""
+    "!pip install \"labelbox[data]==3.0.0rc2\""
    ]
   },
   {
@@ -246,7 +246,7 @@
     "image_url = \"https://picsum.photos/id/1003/200/300\"\n",
     "image = RasterData(url = image_url)\n",
     "# All of these can access the numpy representation the same way:\n",
-    "np_data = image.data\n",
+    "np_data = image.value\n",
     "\n",
     "\n",
     "im = Image.fromarray(np_data)\n",
@@ -273,7 +273,7 @@
     "text = TextData(text = \" some text content...\")\n",
     "text = TextData(url = \"https://filesamples.com/samples/document/txt/sample3.txt\")\n",
     "\n",
-    "print(text.data[:100])"
+    "print(text.value[:100])"
    ]
   },
   {
@@ -306,7 +306,7 @@
    "outputs": [],
    "source": [
     "\n",
-    "for idx, frame in video.data:\n",
+    "for idx, frame in video.value:\n",
     "    # Show every 50th frame\n",
     "    if idx % 50 == 0:\n",
     "        video_im = Image.fromarray(frame)\n",

examples/annotation_types/converters.ipynb

@@ -25,7 +25,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install \"labelbox[data]==3.0.0rc1\""
+    "!pip install \"labelbox[data]==3.0.0rc2\""
    ]
   },
   {
@@ -361,7 +361,7 @@
     "\n",
     "for label in labels:\n",
     "    annotation_lookup = label.frame_annotations()\n",
-    "    for idx, frame in label.data.data:\n",
+    "    for idx, frame in label.data.value:\n",
     "        if idx % 30 != 1:\n",
     "            continue\n",
     "            \n",
@@ -412,7 +412,7 @@
     }
    ],
    "source": [
-    "im_data = label_list[0].data.data\n",
+    "im_data = label_list[0].data.value\n",
     "h,w = im_data.shape[:2]\n",
     "Image.fromarray(im_data)"
    ]

examples/annotation_types/label_containers.ipynb

@@ -20,7 +20,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install \"labelbox[data]==3.0.0rc1\""
+    "!pip install \"labelbox[data]==3.0.0rc2\""
    ]
   },
   {

examples/annotation_types/mal_using_annotation_types.ipynb

@@ -17,7 +17,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install \"labelbox[data]==3.0.0rc1\""
+    "!pip install \"labelbox[data]==3.0.0rc2\""
    ]
   },
   {
@@ -203,7 +203,7 @@
     "labellist = LabelList()\n",
     "for image_url in image_urls:\n",
     "    image_data = RasterData(url = image_url)\n",
-    "    height, width = image_data.data.shape[:2]\n",
+    "    height, width = image_data.value.shape[:2]\n",
     "    prediction = predict(np.array([image_data.im_bytes]), min_score=0.5, height=height, width = width)\n",
     "    annotations = get_annotations(prediction['boxes'], prediction['class_indices'], prediction['seg_masks'])\n",
     "    update_bag_classifications(annotations)\n",

examples/model_diagnostics/model_diagnostics_demo.ipynb

@@ -73,7 +73,7 @@
    },
    "outputs": [],
    "source": [
-    "!pip install \"labelbox[data]==3.0.0rc1\" \\\n",
+    "!pip install \"labelbox[data]==3.0.0rc2\" \\\n",
     "             scikit-image \\\n",
     "             tensorflow"
    ]
@@ -254,7 +254,7 @@
     "predictions = LabelList()\n",
     "for (image_url, external_id) in notebook.tqdm(image_data):\n",
     "    image = RasterData(url = image_url, external_id = external_id)\n",
-    "    height, width = image.data.shape[:2]\n",
+    "    height, width = image.value.shape[:2]\n",
     "    prediction = predict(np.array([image.im_bytes]), min_score=0.5, height=height, width = width)\n",
     "    boxes, classes, seg_masks = prediction[\"boxes\"], prediction[\"class_indices\"], prediction[\"seg_masks\"]\n",
     "    annotations = []\n",

examples/model_diagnostics/model_diagnostics_guide.ipynb

@@ -88,7 +88,7 @@
    },
    "outputs": [],
    "source": [
-    "!pip install \"labelbox[data]==3.0.0rc1\""
+    "!pip install \"labelbox[data]==3.0.0rc2\""
    ]
   },
   {
@@ -258,7 +258,7 @@
     "    \n",
     "    # Build a list of annotation objects from your model inferences\n",
     "    # image.data is just the numpy array representation of the image\n",
-    "    prediction = model.predict(image.data)\n",
+    "    prediction = model.predict(image.value)\n",
     "    # Iterate of segmentation channels or instances depending on your model architecture ( or both )\n",
     "    for instance, (xmin, ymin, xmax, ymax), seg, class_idx in prediction:\n",
     "        class_name = class_names.get(class_idx)\n",

labelbox/data/annotation_types/data/raster.py

@@ -59,7 +59,7 @@ def np_to_bytes(self, arr: np.ndarray) -> bytes:
         return im_bytes.getvalue()
 
     @property
-    def data(self) -> np.ndarray:
+    def value(self) -> np.ndarray:
         """
         Property that unifies the data access pattern for all references to the raster.
 

labelbox/data/annotation_types/data/text.py

@@ -16,7 +16,7 @@ class TextData(BaseData):
     url: Optional[str] = None
 
     @property
-    def data(self) -> str:
+    def value(self) -> str:
         """
         Property that unifies the data access pattern for all references to the text.
 

labelbox/data/annotation_types/data/video.py

@@ -41,7 +41,7 @@ def load_frames(self, overwrite: bool = False) -> None:
             self.frames[count] = frame
 
     @property
-    def data(self):
+    def value(self):
         return self.frame_generator()
 
     def frame_generator(

tests/data/annotation_types/data/test_raster.py

@@ -19,21 +19,21 @@ def test_im_bytes():
     im_bytes = BytesIO()
     Image.fromarray(data).save(im_bytes, format="PNG")
     raster_data = RasterData(im_bytes=im_bytes.getvalue())
-    data_ = raster_data.data
+    data_ = raster_data.value
     assert np.all(data == data_)
 
 
 def test_im_url():
     raster_data = RasterData(url="https://picsum.photos/id/829/200/300")
-    data_ = raster_data.data
+    data_ = raster_data.value
     assert data_.shape == (300, 200, 3)
 
 
 def test_im_path():
     img_path = "/tmp/img.jpg"
     urllib.request.urlretrieve("https://picsum.photos/id/829/200/300", img_path)
     raster_data = RasterData(file_path=img_path)
-    data_ = raster_data.data
+    data_ = raster_data.value
     assert data_.shape == (300, 200, 3)