scitex 2.13.0__py3-none-any.whl → 2.15.1__py3-none-any.whl

scitex/scholar/docs/to_claude/examples/example-python-project-scitex/scripts/mnist/plot_digits.py

@@ -1,118 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Timestamp: "2025-02-15 01:56:49 (ywatanabe)"
-# File: /home/ywatanabe/proj/example-scitex-project/scripts/mnist/plot_digits.py
-
-__file__ = "./scripts/mnist/plot_digits.py"
-
-"""
-Functionality:
-    - Visualizes MNIST dataset samples
-Input:
-    - MNIST dataset
-Output:
-    - Sample image plots
-Prerequisites:
-    - scitex package
-    - PyTorch
-"""
-
-"""Imports"""
-import argparse
-from typing import Optional
-
-import matplotlib.pyplot as plt
-import scitex
-from torch.utils.data import DataLoader
-
-"""Parameters"""
-
-"""Functions & Classes"""
-
-
-def plot_samples(loader: DataLoader, n_samples: int = 25) -> None:
-    images, labels = next(iter(loader))
-    fig, axes = scitex.plt.subplots(5, 5, figsize=(10, 10))
-
-    for idx, ax in enumerate(axes.flat):
-        if idx < n_samples:
-            ax.imshow(images[idx].squeeze(), cmap="gray")
-            ax.set_title(f"Label: {labels[idx]}")
-            # ax.axis("off")
-
-    plt.tight_layout()
-    scitex.io.save(fig, CONFIG.PATH.MNIST.FIGURES + "mnist_samples.jpg", symlink_from_cwd=True)
-
-
-def plot_label_examples(loader: DataLoader) -> None:
-    images, labels = next(iter(loader))
-    fig, axes = scitex.plt.subplots(2, 5, figsize=(15, 6))
-
-    label_examples = {}
-    for img, label in zip(images, labels):
-        if label.item() not in label_examples and len(label_examples) < 10:
-            label_examples[label.item()] = img
-
-    for idx, (label, img) in enumerate(sorted(label_examples.items())):
-        row, col = idx // 5, idx % 5
-        axes[row, col].imshow(img.squeeze(), cmap="gray")
-        axes[row, col].set_title(f"Digit: {label}")
-        # axes[row, col].axis("off")
-
-    plt.tight_layout()
-    scitex.io.save(fig, CONFIG.PATH.MNIST.FIGURES + "mnist_digits.jpg", symlink_from_cwd=True)
-
-
-def main(args: argparse.Namespace) -> Optional[int]:
-    train_loader = scitex.io.load(CONFIG.PATH.MNIST.LOADER.TRAIN)
-    plot_samples(train_loader)
-    plot_label_examples(train_loader)
-    return 0
-
-
-def parse_args() -> argparse.Namespace:
-    parser = argparse.ArgumentParser(description="Visualize MNIST samples")
-    args = parser.parse_args()
-    scitex.str.printc(args, c="yellow")
-    return args
-
-
-def run_main() -> None:
-    """Initialize scitex framework, run main function, and cleanup.
-
-    scitex framework manages:
-      - Parameters defined in yaml files under `./config dir`
-      - Setting saving directory (/path/to/file.py -> /path/to/file.py_out/)
-      - Symlink for `./data` directory
-      - Logging timestamp, stdout, stderr, and parameters
-      - Matplotlib configurations (also, `scitex.plt` will track plotting data)
-      - Random seeds
-
-    THUS, DO NOT MODIFY THIS RUN_MAIN FUNCTION
-    """
-    import sys
-
-    import matplotlib.pyplot as plt
-
-    global CONFIG, CC, sys, plt
-    args = parse_args()
-    CONFIG, sys.stdout, sys.stderr, plt, CC = scitex.gen.start(
-        sys,
-        plt,
-        args=args,
-        file=__file__,
-        agg=True,
-    )
-
-    exit_status = main(args)
-
-    scitex.gen.close(
-        CONFIG,
-        exit_status=exit_status,
-    )
-
-
-if __name__ == "__main__":
-    run_main()
-
-# EOF

scitex/scholar/docs/to_claude/examples/example-python-project-scitex/scripts/mnist/plot_umap_space.py

@@ -1,110 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Timestamp: "2025-02-15 01:06:57 (ywatanabe)"
-# File: /home/ywatanabe/proj/example-scitex-project/scripts/mnist/plot_umap_space.py
-
-__file__ = "./scripts/mnist/plot_umap_space.py"
-
-"""
-Functionality:
-    - Creates UMAP visualization of MNIST dataset
-Input:
-    - MNIST dataset
-Output:
-    - UMAP visualization plots
-Prerequisites:
-    - scitex package
-    - umap-learn
-"""
-
-"""Imports"""
-import argparse
-from typing import Optional
-
-import matplotlib.pyplot as plt
-import scitex
-import numpy as np
-import umap
-
-"""Parameters"""
-
-
-"""Functions & Classes"""
-
-
-def create_umap_embedding(data: np.ndarray) -> np.ndarray:
-    reducer = umap.UMAP(random_state=CONFIG.MNIST.UMAP_RANDOM_STATE, n_jobs=-1)
-    embedding = reducer.fit_transform(data)
-    return embedding
-
-
-def plot_umap(embedding: np.ndarray, labels: np.ndarray) -> None:
-    fig, ax = scitex.plt.subplots(figsize=(12, 8))
-    scatter = ax.scatter(
-        embedding[:, 0], embedding[:, 1], c=labels, cmap="tab10", alpha=0.5
-    )
-
-    plt.colorbar(scatter)
-    ax.set_title("UMAP Projection of MNIST Digits")
-    ax.set_xlabel("UMAP 1")
-    ax.set_ylabel("UMAP 2")
-
-    scitex.io.save(fig, CONFIG.PATH.MNIST.FIGURES + "umap.jpg", symlink_from_cwd=True)
-
-
-def main(args: argparse.Namespace) -> Optional[int]:
-    train_data = scitex.io.load(CONFIG.PATH.MNIST.FLATTENED.TRAIN)
-    train_labels = scitex.io.load(CONFIG.PATH.MNIST.LABELS.TRAIN)
-    embedding = create_umap_embedding(train_data)
-    plot_umap(embedding, train_labels)
-    return 0
-
-
-def parse_args() -> argparse.Namespace:
-    parser = argparse.ArgumentParser(
-        description="Create UMAP visualization of MNIST"
-    )
-    args = parser.parse_args()
-    scitex.str.printc(args, c="yellow")
-    return args
-
-
-def run_main() -> None:
-    """Initialize scitex framework, run main function, and cleanup.
-
-    scitex framework manages:
-      - Parameters defined in yaml files under `./config dir`
-      - Setting saving directory (/path/to/file.py -> /path/to/file.py_out/)
-      - Symlink for `./data` directory
-      - Logging timestamp, stdout, stderr, and parameters
-      - Matplotlib configurations (also, `scitex.plt` will track plotting data)
-      - Random seeds
-
-    THUS, DO NOT MODIFY THIS RUN_MAIN FUNCTION
-    """
-    import sys
-
-    import matplotlib.pyplot as plt
-
-    global CONFIG, CC, sys, plt
-    args = parse_args()
-    CONFIG, sys.stdout, sys.stderr, plt, CC = scitex.gen.start(
-        sys,
-        plt,
-        args=args,
-        file=__file__,
-        agg=True,
-    )
-
-    exit_status = main(args)
-
-    scitex.gen.close(
-        CONFIG,
-        exit_status=exit_status,
-    )
-
-
-if __name__ == "__main__":
-    run_main()
-
-# EOF

scitex/scholar/docs/to_claude/guidelines/README.md

@@ -1,9 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-05-17 04:54:09
-!-- Author: ywatanabe
-!-- File: /ssh:ywatanabe@sp:/home/ywatanabe/.dotfiles/.claude/guidelines/README.md
-!-- --- -->
-
-This directory contains guidelines for agents to understand user's intentions.
-
-<!-- EOF -->

scitex/scholar/docs/to_claude/guidelines/USER_PHILOSOPHY/01_DEVELOPMENT_CYCLE.md

@@ -1,47 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-09-01 08:14:57
-!-- Author: ywatanabe
-!-- File: /home/ywatanabe/.dotfiles/.claude/to_claude/guidelines/USER_PHILOSOPHY/01_DEVELOPMENT_CYCLE.md
-!-- --- -->
-
-# Key Principles
-
--   Tests based on **specifications and acceptance criteria**, not just architecture
--   **Quality gates** beyond coverage (linting, type checking)
--   **Git strategy** with `--fixup` commits during red phase
--   **Code quality checks** before commits
--   **Documentation updates** when APIs change
-
-
-# Development Cycle Steps
-
-01. Accomplish tree-like architecture agreement with behavioral specifications
-02. Plan modular cycles with descriptive phase allocation (e.g., Phase 01: Config Development)
-03. Create feature branch from develop branch
-04. Implement red tests based on **specifications and acceptance criteria** for the scope
-05. Verify tests fail for the right reasons (Red Testing)
-06. Implement source code to satisfy specifications (understanding intent)
-07. Tests with pytest-testmon (only for affected tests from source code change)
-08. Run ruff check/format to ensure code quality before commit
-09. Git commit with appropriate strategy (`--fixup` while red, logical when green)
-10. Repeat 04 -> 09 until quality gates passed:
-    -   All tests passing
-    -   Coverage meets threshold (not always 100%)
-    -   No linting errors
-    -   Type checking passes
-11. Squash git commits if cluttered (`rebase -i --autosquash`)
-12. Run full test suite and quality validation
-13. Update documentation if APIs changed
-14. Merge back to develop branch
-15. Repeat 03 -> 14 until project completion
-
-
-# Quality Gates
-
--   All tests passing
--   Coverage meets threshold
--   No linting errors (ruff)
--   Type checking passes (mypy)
--   Documentation updated for API changes
-
-<!-- EOF -->

scitex/scholar/docs/to_claude/guidelines/USER_PHILOSOPHY/02_NAMING_CONVENSIONS.md

@@ -1,91 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-09-01 08:15:57
-!-- Author: ywatanabe
-!-- File: /home/ywatanabe/.dotfiles/.claude/to_claude/guidelines/USER_PHILOSOPHY/02_NAMING_CONVENSIONS.md
-!-- --- -->
-
-# Naming Conventions
-
-## 1. Explicit is better than implicit
-
--   `mcp_server.py` instead of `server.py` - more specific
--   `ChromaVectorDBManager` instead of generic `VectorStore`
-
-## 2. Methods and functions written in verb_objects format
-
--   Clean MCP tool names: `search`, `get_topics`, `get_database_overview`, `get_file_content`
-
-
-## 3. Classes
-
--   **One class per file**:
-    `_MarkdownChunker.py includes solely `class MarkdownChunker(...)`, with `main()` and main guard
-    `$ python -m package.domain._ClassName.py` will explain the usage
-
--   **File names for Private Classes**:
-    Starting with underscore (`_MarkdownChunker.py`)
-
--   **File names for Public Classes**:
-    Starting without underscore (`MarkdownChunker.py`)
-
--   **Class Names**:
-    Always PascalCase (e.g., `class MarkdownChunker(...)`)
-
--   **Four types of Classes**
-    1.  **Abstract bases**
-        File name with `_Base` and class (`_BaseCar.py` → `class BaseCar`)
-    
-    2.  **Entity**
-        Simple noun (e.g., Car)
-    
-    3.  **Business Logic**
-        Often has suffix `-er` (e.g., CarMaintainer, CarDesigner)
-    
-    4.  **Data Container**
-        File name with underscore and suffix with `Container`
-        (e.g., `_CarSpecsContainer.py` includes @dataclass class CarSpecsContainer(...))
-
--   **Methods**
-    Methods must be in the `verb_objects` format
-
--   **Decorators**
-    Use decorators such as @dataclass, @classmethod, @staticmethod, @abstractmethod, @property
-
-
-## 4. Functions
-
--   **One function per file**
-    `_get_file_type.py` includes solely `def get_file_type(...)`, with `main()` and main guard
-    `$ python -m package.domain._function_file.py` will explain the usage
-
--   **\*Grouped by domain**
-    e.g., `file_utils/_get_file_type.py`, `text_utils/_clean_text.py`
-
--   **File names starting with underscore**
-    e.g., `_get_file_type.py`, `_clean_text.py`, etc.
-
-## 4. Source-Test agreement
-
--   **One-to-one mapping**
-    Each source file maps to the corresponding test file
-    
-    **Mapping step**:
-    Step 1. src -> tests
-    Step 2. filename.py -> test_filename.py
-    
-    Example 1:
-    src:   `/path/to/src/package/domain/filename.py`
-    tests: `/path/to/tests/package/domain/test_filename.py`
-    
-    Note that when source file starts from underscore, use **double underscore** (test_ + _filename.py)
-    
-    Example 2:
-    src:   `/path/to/src/package/domain/_filename.py`
-    tests: `/path/to/tests/package/domain/test__filename.py`
-
-## 4. Split Type Definitions
-
--   **Separate enum files**: `_DocumentType.py`, `_ChunkingStrategy.py`, etc.
--   **Utility functions**: `_DocumentTypeUtils.py` for related functions
-
-<!-- EOF -->

scitex/scholar/docs/to_claude/guidelines/USER_PHILOSOPHY/03_ARCHITECTUAL_AGREEMENT_PROCESS.md

@@ -1,86 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-09-01 08:17:28
-!-- Author: ywatanabe
-!-- File: /home/ywatanabe/.dotfiles/.claude/to_claude/guidelines/USER_PHILOSOPHY/03_ARCHITECTUAL_AGREEMENT_PROCESS.md
-!-- --- -->
-
-
-# AI-Agent Architectural Agreement Process
-
-## Phase 1: Large Picture Agreement
-
-1.  **Examine reference implementations**
-    e.g., example project
-
-2.  **Present tree-like project structure**
-    With full paths and class hierarchies
-    decorators, class name, function name, arguments and returns with type hints
-
-3.  **Get user feedback**
-    Verify user agrees project organization and naming patterns
-    Until agreement reached, `mgmt/ARCHITECTURE_v01.org` file is versioned by suffix
-    e.g., `mgmt/ARCHITECTURE_v02.org`, `mgmt/ARCHITECTURE_v03.org`, ...;
-
-4.  **Iterate based on preferences**
-    Until structure feels "right"
-
-
-## Phase 2: Detailed API Design
-
-1.  **Expand tree structure** with method signatures and arguments
-2.  **Show complete type information** for all functions and classes
-3.  **Present workflow diagrams** showing data flow and relationships
-4.  **Get approval** before moving to implementation
-
-
-## Phase 3: Data Architecture
-
-1.  **Design dataclass system** with clear responsibilities
-    Use `@dataclass` decorator and `Container` suffixed class name
-2.  **Show integration patterns** between services and data objects
-3.  **Document validation and serialization** approaches using Mermaid diagrams
-
-
-# Reusable Template for Future Projects
-
-## Step 1: Discovery Phase
-
-    Questions to ask:
-    1. "Can you show me similar projects you've worked on for reference patterns?"
-    2. "What existing codebases should I examine for conventions?"
-    3. "Are there specific frameworks or patterns you prefer?"
-    4. "How do you typically organize domain logic vs utilities?"
-
-## Step 2: Structure Proposal
-
-    Present:
-    1. Complete directory tree with file naming patterns
-    2. Class hierarchy with public/private distinctions  
-    3. Utility organization (atomic vs grouped)
-    4. Test and example structure mirroring
-
-## Step 3: Naming Convention Agreement
-
-    Establish:
-    1. Service class naming patterns (-er, Manager, Engine, etc.)
-    2. Data object suffixes (Data, Container, Model, etc.)
-    3. Entity naming (clean, no suffixes)
-    4. File naming (private files, public classes)
-
-## Step 4: API Design Review
-
-    Document:
-    1. All classes with method signatures
-    2. Complete type annotations
-    3. Configuration patterns
-    4. Integration examples
-
-## Step 5: Visual Architecture
-
-    Create:
-    1. Workflow diagrams showing data flow
-    2. Class relationship diagrams
-    3. Service integration patterns
-    4. Usage examples and patterns
-
-<!-- EOF -->

scitex/scholar/docs/to_claude/guidelines/USER_PHILOSOPHY/04_ARCHITECTURE_PREDEFINED.md

@@ -1,77 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-09-01 08:20:12
-!-- Author: ywatanabe
-!-- File: /home/ywatanabe/.dotfiles/.claude/to_claude/guidelines/USER_PHILOSOPHY/04_ARCHITECTURE_PREDEFINED.md
-!-- --- -->
-
-
-# Agreed, Predefined Architecture
-
-# External Data Storage
-
-## Confiurations
--   `config/` - Configuration YAML files
-
-## Project Data
--   `data/` - Project data files (datasets, sample data); gitignored
-
-## Documentaitons
--   `docs/` - Documentations for users in general
--   `docs/to_claude` - Guidelines for Claude Code
--   `docs/from_user` - User messages
--   `docs/from_agents` - Agents can write any documents here
-
-## Minimal Experiments
--   `.dev/` - Development sandbox for miniature tests; gitignored
--   But not included in final product
--   Need to incorporate successful code into `src` for permanent use
-
-## Examples
--   `examples/` - For most important usages in the project
--   Use `ipynb` is a better practice for rendering in GitHub
--   But the user is not a notebook person but more of a person with `.py` scripts
-
-## Makefile
--   install, test-changed, test-full, coverage-html, ci-container, ci-act, ci-local,
-    lint, clean, build, upload-pypi-test, upload-pypi, release, and so on
-
-## Project Management Derectory
--   `mgmt` for searchability
--   `mgmt/PROJECT_DESCRIPTION_v01.md`
--   `mgmt/ARCHITECTURE_v01.md`
--   `mgmt/BULLETIN_BOARD_v01.md`
-
-## pyproject.toml
-No need to conventional files such as `setup.py`, `requirements.txt`, `MANIFEST.in`
-
-## Test code
-
-**Test files for source code**
-
-- For pip package:
-    -   `tests/pip_package_name/test_public_function_name.py`
-    -   `tests/pip_package_name/test__private_function_name.py`
-    -   `tests/pip_package_name/test_PublicClassName.py`
-    -   `tests/pip_package_name/test__PrivateClassName.py`
-
-- For scientific projects:
-    -   `tests/scripts/test_public_function_name.py`
-    -   `tests/scripts/test__private_function_name.py`
-    -   `tests/scripts/test_PublicClassName.py`
-    -   `tests/scripts/test__PrivateClassName.py`
-
-**Test files except for source code**
-
--   `tests/custom/`
-
-**Scripts for running GitHub actions locally**
-
--   `tests/github_actions/`
-
-**Test Results Summary**
-
--   `tests/reports/` - Natural location for test outputs
-
-#+END_SRC
-
-<!-- EOF -->

scitex/scholar/docs/to_claude/guidelines/USER_PHILOSOPHY/05_PRIORITY_CONFIG.md

@@ -1,22 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-09-01 08:23:25
-!-- Author: ywatanabe
-!-- File: /home/ywatanabe/.dotfiles/.claude/to_claude/guidelines/USER_PHILOSOPHY/05_PRIORITY_CONFIG.md
-!-- --- -->
-
-# Centraliazed, PriorityConfig
-
--   **./config**
-    Centralized directory for parameters, meaning scripts should avoid to define parameters as much as possible
-
--   **Logically Separated YAML files**
-    e.g., `./config/PATH.yaml`, `./config/COLORS.yaml`
-
--   **Precedence with Priority**
-    direct → config → env → default
-
--   **Type conversion**: Automatic type casting
-
-Please see the official documentation (`./docs/to_claude/guidelines/python/official/priority_config-README.md`)
-
-<!-- EOF -->

scitex/scholar/docs/to_claude/guidelines/USER_PHILOSOPHY/06_MULTIPLE_SPECIAL_AGENTS.md

@@ -1,35 +0,0 @@
-<!-- ---
-!-- Timestamp: 2025-09-01 08:35:19
-!-- Author: ywatanabe
-!-- File: /home/ywatanabe/.dotfiles/.claude/to_claude/guidelines/USER_PHILOSOPHY/06_MULTIPLE_SPECIAL_AGENTS.md
-!-- --- -->
-
-# Agent Overview
-
-Multi-agent system can be beneficial for asynchronous work, context preservation, specialized expertise, reusability, and flexible permissions.
-
-## Trust other agents
-All agents have their special knowledge and skills so please delegate related tasks to other specialists as much as possible. So, when you receive a task, you need to check if there is more appropriate agent on behalf of you.
-
-## No hesitation for concurrent work
-One of the upside of multi agent system is the asynchronous work to execute tasks in parallel. It would be often the case that multiple agents are called, even when they have the exact same role, being expected to work collaborately with minimal overlaps. For example, when test code needs to improve, Tester Agents will be called at the same time.
-
-## Communication Channel
-We only rely on the bulletin board (`./mgmt/BULLETIN_BOARD.md`) for agent-to-agent communication. Please read others entries and write your progress to share momentum as a team.
-
-
-## Agent List
-You can see all the definitions of agents in `.claude/agents/*.md`
-
-BulletinBoardOrganizerAgent.md
-DebuggerAgent.md
-ExamplesDeveloperAgent.md
-ExperimentationAgent.md
-GitHandlerAgent.md
-SciTeXTranslatorAgent.md
-SourceDeveloperAgent.md
-TestDeveloperAgent.md
-TestResultsReportAgent.md
-TestRunnerAgent.md
-
-<!-- EOF -->