torchquery 1.0.2__tar.gz

torchquery-1.0.2/PKG-INFO

@@ -0,0 +1,16 @@
+Metadata-Version: 2.4
+Name: torchquery
+Version: 1.0.2
+Summary: Billion-scale neural healing engine for high-speed tensor manipulation.
+Author: Sundaram Gupta
+License: MIT
+Keywords: torch,pytorch,tensor,deep-learning
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+Requires-Dist: torch>=2.0.0
+
+## 🔌 Compatibility
+**TorchQuery** is built specifically for the **PyTorch** ecosystem. It supports all `torch.Tensor` types and is optimized for both CPU and NVIDIA CUDA (GPU) workflows. It serves as a high-speed alternative to standard **torch** data cleaning methods.

torchquery-1.0.2/README.md

@@ -0,0 +1,2 @@
+## 🔌 Compatibility
+**TorchQuery** is built specifically for the **PyTorch** ecosystem. It supports all `torch.Tensor` types and is optimized for both CPU and NVIDIA CUDA (GPU) workflows. It serves as a high-speed alternative to standard **torch** data cleaning methods.

torchquery-1.0.2/pyproject.toml

@@ -0,0 +1,23 @@
+[build-system]
+requires = ["setuptools>=61.0", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "torchquery"
+version = "1.0.2"  # Bump to 1.0.2 for a fresh start
+authors = [
+  { name="Sundaram Gupta" },
+]
+description = "Billion-scale neural healing engine for high-speed tensor manipulation."
+readme = "README.md"
+requires-python = ">=3.8"
+license = {text = "MIT"}
+dependencies = [
+    "torch>=2.0.0",
+]
+keywords = ["torch", "pytorch", "tensor", "deep-learning"]
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Operating System :: OS Independent",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+]

torchquery-1.0.2/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

torchquery-1.0.2/setup.py

@@ -0,0 +1,7 @@
+from setuptools import setup
+
+setup(
+    name="torchquery",
+    version="1.0.1",
+    install_requires=["torch>=2.0.0"],
+)

torchquery-1.0.2/torchquery.egg-info/PKG-INFO

@@ -0,0 +1,16 @@
+Metadata-Version: 2.4
+Name: torchquery
+Version: 1.0.2
+Summary: Billion-scale neural healing engine for high-speed tensor manipulation.
+Author: Sundaram Gupta
+License: MIT
+Keywords: torch,pytorch,tensor,deep-learning
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+Requires-Dist: torch>=2.0.0
+
+## 🔌 Compatibility
+**TorchQuery** is built specifically for the **PyTorch** ecosystem. It supports all `torch.Tensor` types and is optimized for both CPU and NVIDIA CUDA (GPU) workflows. It serves as a high-speed alternative to standard **torch** data cleaning methods.

torchquery-1.0.2/torchquery.egg-info/SOURCES.txt

@@ -0,0 +1,9 @@
+README.md
+pyproject.toml
+setup.py
+torchquery.py
+torchquery.egg-info/PKG-INFO
+torchquery.egg-info/SOURCES.txt
+torchquery.egg-info/dependency_links.txt
+torchquery.egg-info/requires.txt
+torchquery.egg-info/top_level.txt

torchquery-1.0.2/torchquery.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

torchquery-1.0.2/torchquery.egg-info/requires.txt

@@ -0,0 +1 @@
+torch>=2.0.0

torchquery-1.0.2/torchquery.egg-info/top_level.txt

@@ -0,0 +1 @@
+torchquery

torchquery-1.0.2/torchquery.py

@@ -0,0 +1,74 @@
+import torch
+
+class QueryEngine:
+    def __init__(self, data):
+        """Unified Engine for Mouse Hacker Data Processing."""
+        if torch.is_tensor(data):
+            self.data = data
+        else:
+            self.data = torch.as_tensor(data, dtype=torch.float32)
+
+    @classmethod
+    def pre_load_billions(cls, file_path, count):
+        """Loads billion-scale binary data via Memory Mapping."""
+        storage = torch.FloatStorage.from_file(file_path, shared=True, size=count)
+        return cls(torch.FloatTensor(storage))
+
+    # --- THE HEALERS ---
+    def find_Nan_FillwithModernNumbers(self, modern_value=0.0):
+        """Replaces NaNs with a specific value."""
+        self.data.masked_fill_(torch.isnan(self.data), modern_value)
+        return self.data
+
+    def fill_with_Nan_into_your_suggestions(self):
+        """
+        AI SUGGESTION: Automatically fills NaNs with the Mean (Average).
+        Essential for stable Deep Learning in Tensorkite.
+        """
+        clean_data = self.data[~torch.isnan(self.data)]
+        suggestion = torch.mean(clean_data) if clean_data.numel() > 0 else torch.tensor(0.0)
+        return self.find_Nan_FillwithModernNumbers(modern_value=suggestion.item())
+
+    def find_inf_rename(self, replace_with=0.0):
+        """Cleans Infinity values using in-place operations."""
+        torch.nan_to_num(self.data, posinf=replace_with, neginf=-replace_with, out=self.data)
+        return self.data
+
+    def find_inf_intoleastNum(self):
+        """Replaces Infinity with the minimum value found in the data."""
+        least = torch.min(self.data)
+        return self.find_inf_rename(replace_with=least.item())
+
+    def find_inf_intoBignumbers(self):
+        """Replaces Infinity with the maximum value found in the data."""
+        large = torch.max(self.data)
+        return self.find_inf_rename(replace_with=large.item())
+
+    # --- THE ANALYSTS ---
+    def find_least_rename_largerNumber(self):
+        """Swaps the smallest values with the largest."""
+        least, large = torch.min(self.data), torch.max(self.data)
+        self.data[self.data == least] = large
+        return self.data
+
+    def find_largennum_leastnumbers(self):
+        """Swaps the largest values with the smallest."""
+        least, large = torch.min(self.data), torch.max(self.data)
+        self.data[self.data == large] = least
+        return self.data
+
+class TensorTricafig:
+    def __init__(self, tensor_data):
+        self.data = tensor_data
+
+    def action_normalize(self):
+        """Scales tensor values to a range of 0.0 to 1.0."""
+        min_v, max_v = torch.min(self.data), torch.max(self.data)
+        diff = max_v - min_v
+        if diff != 0:
+            self.data.sub_(min_v).div_(diff)
+        return self.data
+
+    def tensor_grid(self, rows, cols):
+        """Shapes data into a matrix view without copying memory."""
+        return self.data.view(rows, cols)