v 2.0.0

__init__.py

@@ -1,8 +1,7 @@
-from .src import ImageSetAreaNode, FloatImageCombineNode, XYPlotNode
+from .src import LatentCombineNode, XYPlotNode
 
 
 NODE_CLASS_MAPPINGS = {
-    "ImageSetArea": ImageSetAreaNode,
-    "FloatImageCombine": FloatImageCombineNode,
     "XYPlot": XYPlotNode,
+    "LatentCombine": LatentCombineNode,
 }

src/__init__.py

@@ -1,3 +1,2 @@
-from .nodes.image_set_area import ImageSetAreaNode
-from .nodes.float_image_combine import FloatImageCombineNode
 from .nodes.xy_plot import XYPlotNode
+from .nodes.latent_combine import LatentCombineNode

src/base.py

@@ -1,4 +1,5 @@
 import typing as t
+from dataclasses import dataclass
 
 
 class BasePlotNode():
@@ -6,5 +7,10 @@ class BasePlotNode():
     FUNCTION: str = "execute"
 
 
+@dataclass
+class KSamplerXYPlotInput():
+    setting: str
+    value: int
+
+
 Image = t.Any
-FloatImage = list[Image]

src/nodes/float_image_combine.py

@@ -1,19 +1,23 @@
 import typing as t
 
-from ..base import BasePlotNode, FloatImage
+from ..base import BasePlotNode, Image
 
 
 class FloatImageCombineNode(BasePlotNode):
-    RETURN_TYPES: t.Tuple[str] = ("FLOAT_IMAGE",)
+    RETURN_TYPES: t.Tuple[str] = ("IMAGES",)
 
     @classmethod
     def INPUT_TYPES(cls) -> t.Dict[str, t.Any]:
         return {
             "required": {
-                "float_image_1": ("FLOAT_IMAGE",),
-                "float_image_2": ("FLOAT_IMAGE",),
+                "float_image_1": ("IMAGES",),
+                "float_image_2": ("IMAGES",),
             },
         }
 
-    def execute(self, float_image_1: FloatImage, float_image_2: FloatImage) -> tuple[FloatImage]:
+    def execute(
+        self,
+        float_image_1: t.List[Image],
+        float_image_2: t.List[Image],
+    ) -> t.Tuple[t.List[Image]]:
         return (float_image_1 + float_image_2,)

src/nodes/image_set_area.py

@@ -1,13 +1,10 @@
 import typing as t
 
-from ..base import BasePlotNode, FloatImage, Image
+from ..base import BasePlotNode, Image
 
 
 class ImageSetAreaNode(BasePlotNode):
-    RETURN_TYPES: t.Tuple[str] = ("FLOAT_IMAGE",)
-
-    def __init__(self):
-        pass
+    RETURN_TYPES: t.Tuple[str] = ("IMAGES",)
 
     @classmethod
     def INPUT_TYPES(cls) -> t.Dict[str, t.Any]:
@@ -17,5 +14,5 @@ class ImageSetAreaNode(BasePlotNode):
             },
         }
 
-    def execute(self, image: Image) -> tuple[FloatImage]:
+    def execute(self, image: Image) -> t.Tuple[t.List[Image]]:
         return ([image],)

src/nodes/k_sampler_xy_plot.py

@@ -0,0 +1,62 @@
+import typing as t
+
+from nodes import KSamplerAdvanced  # type: ignore
+
+from ..base import BasePlotNode, Image, KSamplerXYPlotInput
+
+
+class KSamplerXYPlotNode(BasePlotNode):
+    RETURN_TYPES: t.Tuple[str] = ("IMAGES",)
+
+    def __init__(self) -> None:
+        self._sampler = KSamplerAdvanced()
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        result = KSamplerAdvanced.INPUT_TYPES()
+        result["required"]["vae"] = ("VAE", )
+        #result["required"]["x_items"] = ("XYPlotItem",)
+        #result["required"]["y_items"] = ("XYPlotItem",)
+        return result
+
+    def execute(
+        self,
+        vae,
+        #x_items,
+        #y_items,
+        **sampler_kw,
+    ) -> tuple[t.List[Image]]:
+        x_items = [
+            KSamplerXYPlotInput(value=1, setting="cfg"),
+            KSamplerXYPlotInput(value=2, setting="cfg"),
+        ]
+        y_items = [
+            KSamplerXYPlotInput(value=1, setting="noise_seed"),
+            KSamplerXYPlotInput(value=2, setting="noise_seed"),
+        ]
+
+        latents = self._sample_latents(
+            x_items=x_items,
+            y_items=y_items,
+            sampler_kw=sampler_kw,
+        )
+        result = list(self._decode_latents(latents=latents, vae=vae))
+        print(result)
+        print(type(result[0]))
+
+        return (result,)
+
+    def _sample_latents(self, x_items, y_items, sampler_kw):
+        for x in x_items:
+            for y in y_items:
+                sampler_settings = sampler_kw.copy()
+                sampler_settings[x.setting] = x.value
+                sampler_settings[y.setting] = y.value
+
+                yield self._sampler.sample(**sampler_settings)[0]
+
+    def _decode_latents(self, latents, vae) -> t.Iterable[Image]:
+        return (
+            vae.decode(i["samples"])
+            for i in latents
+        )

src/nodes/latent_combine.py

@@ -0,0 +1,29 @@
+import typing as t
+
+import torch
+
+from ..base import BasePlotNode, Image
+
+
+class LatentCombineNode(BasePlotNode):
+    RETURN_TYPES: t.Tuple[str] = ("LATENT",)
+
+    @classmethod
+    def INPUT_TYPES(cls) -> t.Dict[str, t.Any]:
+        return {
+            "required": {
+                "latent_1": ("LATENT",),
+                "latent_2": ("LATENT",),
+            },
+        }
+
+    def execute(
+        self,
+        latent_1: t.Dict[str, t.Any],
+        latent_2: t.Dict[str, t.Any],
+    ) -> t.Tuple[t.Dict[str, t.Any]]:
+        latent_1_samples = latent_1["samples"]
+        latent_2_samples = latent_2["samples"]
+        samples = torch.cat((latent_1_samples, latent_2_samples), 0)
+
+        return ({"samples": samples},)

src/nodes/xy_plot.py

@@ -1,6 +1,6 @@
 import typing as t
 
-from ..base import BasePlotNode, FloatImage, Image
+from ..base import BasePlotNode, Image
 from ..utils import tensor_to_pillow, pillow_to_tensor, create_image_grid
 
 
@@ -11,7 +11,7 @@ class XYPlotNode(BasePlotNode):
     def INPUT_TYPES(cls) -> t.Dict[str, t.Any]:
         return {
             "required": {
-                "float_image": ("FLOAT_IMAGE",),
+                "images": ("IMAGE",),
                 "gap": ("INT", {"default": 0, "min": 0}),
                 "nrow": ("INT", {"default": 1, "min": 1}),
             },
@@ -19,11 +19,11 @@ class XYPlotNode(BasePlotNode):
 
     def execute(
         self,
-        float_image: FloatImage,
+        images: Image,
         nrow: int,
         gap: int
     ) -> tuple[Image]:
-        pillow_images = [tensor_to_pillow(i) for i in float_image]
+        pillow_images = [tensor_to_pillow(i) for i in images]
         pillow_grid = create_image_grid(pillow_images, nrow=nrow, gap=gap)
         tensor_grid = pillow_to_tensor(pillow_grid)
 

src/utils.py

@@ -13,35 +13,27 @@ def pillow_to_tensor(image):
     return torch.from_numpy(np.array(image).astype(np.float32) / 255.0).unsqueeze(0)
 
 
-def create_image_grid(images, gap, nrow):
-    """
-    Create a grid of images with a specified gap and number of rows.
+def create_image_grid(images: t.List[Image.Image], gap: int, ncol: int):
+    # Calculate the number of rows needed based on the number of images and columns
+    nrow = (len(images) + ncol - 1) // ncol
 
-    Args:
-        images (List[PIL.Image.Image]): List of images to be placed in the grid.
-        gap (int, optional): Gap between images in pixels. Defaults to 10.
-        nrow (int, optional): Number of rows in the grid. Defaults to 3.
+    # Get the size of the first image to use as a template for the grid
+    size = images[0].size
 
-    Returns:
-        PIL.Image.Image: The merged image grid.
-    """
-    # Calculate number of columns based on number of rows and images
-    ncol = (len(images) + nrow - 1) // nrow
+    # Calculate the total size of the grid with gaps
+    width = size[0] * ncol + gap * (ncol - 1)
+    height = size[1] * nrow + gap * (nrow - 1)
 
-    # Get size of each image in pixels
-    image_width, image_height = images[0].size
+    # Create a new image for the grid
+    grid_image = Image.new("RGB", (width, height), color="white")
 
-    # Create new image to hold the grid
-    grid_width = ncol * image_width + (ncol - 1) * gap
-    grid_height = nrow * image_height + (nrow - 1) * gap
-    grid_image = Image.new("RGB", (grid_width, grid_height), color="white")
-
-    # Paste images into grid
+    # Iterate over each image and paste it into the grid
     for i, image in enumerate(images):
-        row = i // ncol
-        col = i % ncol
-        x = col * (image_width + gap)
-        y = row * (image_height + gap)
+        # Calculate the position of the image in the grid
+        x = (i % ncol) * (size[0] + gap)
+        y = (i // ncol) * (size[1] + gap)
+
+        # Paste the image into the grid
         grid_image.paste(image, (x, y))
 
     return grid_image

workflows/xy_plot_base.json

@@ -131,8 +131,8 @@
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             13
           ],
@@ -637,8 +637,8 @@
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             22
           ],
@@ -672,8 +672,8 @@
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             12
           ],
@@ -701,19 +701,19 @@
       "inputs": [
         {
           "name": "float_image_1",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 12
         },
         {
           "name": "float_image_2",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 13
         }
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             32
           ],
@@ -741,19 +741,19 @@
       "inputs": [
         {
           "name": "float_image_1",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 32
         },
         {
           "name": "float_image_2",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 22
         }
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             24
           ],
@@ -780,8 +780,8 @@
       "mode": 0,
       "inputs": [
         {
-          "name": "float_image",
-          "type": "FLOAT_IMAGE",
+          "name": "images",
+          "type": "IMAGES",
           "link": 24
         }
       ],
@@ -870,7 +870,7 @@
       0,
       12,
       0,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       13,
@@ -878,7 +878,7 @@
       0,
       12,
       1,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       20,
@@ -894,7 +894,7 @@
       0,
       19,
       1,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       24,
@@ -902,7 +902,7 @@
       0,
       17,
       0,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       32,
@@ -910,7 +910,7 @@
       0,
       19,
       0,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       35,
@@ -1133,4 +1133,4 @@
   "config": {},
   "extra": {},
   "version": 0.4
-}
+}

workflows/xy_plot_mini.json

@@ -50,8 +50,8 @@
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             12
           ],
@@ -79,19 +79,19 @@
       "inputs": [
         {
           "name": "float_image_1",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 12
         },
         {
           "name": "float_image_2",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 13
         }
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             32
           ],
@@ -119,19 +119,19 @@
       "inputs": [
         {
           "name": "float_image_1",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 32
         },
         {
           "name": "float_image_2",
-          "type": "FLOAT_IMAGE",
+          "type": "IMAGES",
           "link": 22
         }
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             24
           ],
@@ -166,8 +166,8 @@
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             13
           ],
@@ -201,8 +201,8 @@
       ],
       "outputs": [
         {
-          "name": "FLOAT_IMAGE",
-          "type": "FLOAT_IMAGE",
+          "name": "IMAGES",
+          "type": "IMAGES",
           "links": [
             22
           ],
@@ -303,8 +303,8 @@
       "mode": 0,
       "inputs": [
         {
-          "name": "float_image",
-          "type": "FLOAT_IMAGE",
+          "name": "images",
+          "type": "IMAGES",
           "link": 24
         }
       ],
@@ -371,7 +371,7 @@
       0,
       12,
       0,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       13,
@@ -379,7 +379,7 @@
       0,
       12,
       1,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       20,
@@ -395,7 +395,7 @@
       0,
       19,
       1,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       24,
@@ -403,7 +403,7 @@
       0,
       17,
       0,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       32,
@@ -411,7 +411,7 @@
       0,
       19,
       0,
-      "FLOAT_IMAGE"
+      "IMAGES"
     ],
     [
       73,
@@ -442,4 +442,4 @@
   "config": {},
   "extra": {},
   "version": 0.4
-}
+}