ras-commander 0.61.0__tar.gz → 0.65.0__tar.gz

{ras_commander-0.61.0/ras_commander.egg-info → ras_commander-0.65.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: ras-commander
-Version: 0.61.0
+Version: 0.65.0
 Summary: A Python library for automating HEC-RAS 6.x operations
 Home-page: https://github.com/gpt-cmdr/ras-commander
 Author: William M. Katzenmeyer, P.E., C.F.M.
@@ -35,22 +35,25 @@ Dynamic: summary
 RAS Commander is a Python library for automating HEC-RAS operations, providing a set of tools to interact with HEC-RAS project files, execute simulations, and manage project data. This library was initially conceptualized in the Australian Water School course "AI Tools for Modelling Innovation", and subsequently expanded to cover the basic functionality of the HECRASController COM32 interface using open-source python libraries.  This library uses a Test Driven Development strategy, leveraging the publicly-available HEC-RAS Example projects to create repeatable demonstration examples.  The "Commmander" moniker is inspired by the "Command Line is All You Need" approach to HEC-RAS automation that was first implemented in the HEC-Commander Tools repository. 
 
 ## Repository Author:
-(William Katzenmeyer, P.E., C.F.M.)[https://engineeringwithllms.info]
+[William Katzenmeyer, P.E., C.F.M.](https://engineeringwithllms.info)
+-----  
 
-## Don't Ask Me, Ask GPT! 
+## Don't Ask Me, Ask GPT!
 
 This repository has several methods of interaction with Large Language Models and LLM-Assisted Coding built right in: 
 
-1. **[RAS Commander Library Assistant GPT](https://chatgpt.com/g/g-TZRPR3oAO-ras-commander-library-assistant)**: A specialized GPT model with access to the ras-commander codebase and library, available for answering queries and providing code suggestions.   You can even upload your own plan, unsteady and HDF files to inspect and help determine how to automate your workflows or visualize your results, although this ability is still limited by OpenAI's GPT frameworks and may not be useful for long conversations.
+1. **[Purpose-Built Knowledge Base Summaries](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/llm_knowledge_bases)**: Up-to-date compilations of the documentation and codebase for use with large language models like Claude, ChatGPT, Gemini or Grok. Look in 'ai_tools/assistant_knowledge_bases/' in the repo.  The repo's codebase (without documentation and examples) has been curated to stay within the current ~200k context window limitations of frontier models, and for tasks that do not need an understanding of the underlying code, the Comprehensive Library Guide and any relevant examples from the example folder should be adequate context for leveraging the ras-commander API to complete tasks. 
 
-2. **[Purpose-Built Knowledge Base Summaries](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/assistant_knowledge_bases)**: Up-to-date compilations of the documentation and codebase for use with large language models like Claude or GPT-4. Look in 'ai_tools/assistant_knowledge_bases/' in the repo.  The repo's codebase (without documentation and examples) has been curated to stay within the current ~200k context window limitations of frontier models, and for tasks that do not need an understanding of the underlying code, the Comprehensive Library Guide and any relevant examples from the example folder should be adequate context for leveraging the ras-commander API to complete tasks. 
+2. **[Cursor IDE Integration](https://github.com/gpt-cmdr/ras-commander/blob/main/.cursorrules)**: Custom rules(.cursorrules) for the Cursor IDE to provide context-aware suggestions and documentation.  Just open the repository folder in Cursor to recognize these instructions.  You can create your own folders "/workspace/, "/projects/", or "my_projects/" as these are already in the .gitignore, and place your custom scripts there for your projects.  This will allow easy referencing of the ras-commander documents and individual repo files, the automatic loading of the .cursorrules file.  Alternatvely, download the github repo into your projects folder to easily load documents and use cursor rules files.
 
-3. **[Cursor IDE Integration](https://github.com/gpt-cmdr/ras-commander/blob/main/.cursorrules)**: Custom rules(.cursorrules) for the Cursor IDE to provide context-aware suggestions and documentation.  Just open the repository folder in Cursor to recognize these instructions.  You can create your own folders "/workspace/, "/projects/", or "my_projects/" as these are already in the .gitignore, and place your custom scripts there for your projects.  This will allow easy referencing of the ras-commander documents and individual repo files, the automatic loading of the .cursorrules file.  Alternatvely, download the github repo into your projects folder to easily load documents and use cursor rules files.
+3. **[RAS-Commander Library Assistant](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/library_assistant)**: A full-featured interface for multi-turn conversations, using your own API keys and the ras-commander library for context.  The library assistant allows you to load your own scripts and chat with specific examples and/or function classes in the RAS-Commander library to effectively utilize the library's functions in your workflow.  To reduce hallucinations, a file browser is included which adds full files to the conversation to ensure grounded responses.  A dashboard shows you the total context and estimated cost of each request.  **Now with support for Claude 3.7, OpenAI's o1 and o3-mini, and Deepseek V3 and R1 models using US-based Together.ai**
+
+
+4. **[RAS Commander Library Assistant on ChatGPT](https://chatgpt.com/g/g-TZRPR3oAO-ras-commander-library-assistant)**: A specialized ChatGPT "GPT" with access to the ras-commander codebase and library, available for answering queries and providing code suggestions.   You can even upload your own plan, unsteady and HDF files to inspect and help determine how to automate your workflows or visualize your results.  _NOTE: GPT's are still quite limited by OpenAI's GPT frameworks and may not be useful for long conversations.  Code interpreter cannot run HEC-RAS but can open and view smaller HDF files and projects for demonstration purposes_
 
-5. **[RAS-Commander Library Assistant](https://github.com/gpt-cmdr/ras-commander/blob/main/library_assistant)**: A full-featured interface for multi-turn conversations, using your own API keys and the ras-commander library for context.  The library assistant allows you to load your own scripts and chat with specific examples and/or function classes in the RAS-Commander library to effectively utilize the library's functions in your workflow.  To reduce hallucinations, a file browser is included which adds full files to the conversation to ensure grounded responses.  A dashboard shows you the total context and estimated cost of each request.  **Now with support for OpenAI's o1 and o3-mini, and Deepseek V3 and R1 models using US-based Together.ai**
 
 ## Background
-The ras-commander library emerged from the initial test-bed of AI-driven coding represented by the HEC-Commander tools Python notebooks. These notebooks served as a proof of concept, demonstrating the value proposition of automating HEC-RAS operations. The transition from notebooks to a structured library aims to provide a more robust, maintainable, and extensible solution for water resources engineers.
+The ras-commander library emerged from the initial test-bed of AI-driven coding represented by the [HEC-Commander tools](https://github.com/gpt-cmdr/HEC-Commander) Python notebooks. These notebooks served as a proof of concept, demonstrating the value proposition of automating HEC-RAS operations. In 2024, I taught a series of progressively more complex webinars demonstrating how to use simple prompting, example projects and natural language instruction to effectively code HEC-RAS automation workflows, culminating in a 6 hour course.  The library published for utilization in that course, [awsrastools](https://github.com/gpt-cmdr/awsrastools) served as a foundation of examples which were iteratively extended into the full RAS-Commander library.  Unlike the original notebook by the same name, this library is not focused on parallel execution across multiple machines.  Instead, it is focused on providing a general-purpose python API for interacting with HEC-RAS projects, and building an AI-friendly library that will allow new users to quickly scaffold their own workflows into a python script. Example notebooks are provided, but the intention is to empower engineers, software developers, GIS personnel and data analysts to more easily access and interact with HEC-RAS data in a python environment.  Also, by publishing these examples publicly, with complete working code examples and LLM optimization, future users can readily rewrite they key functions of the library for inclusion in into their own preferred libraries, languages or return formats.
 
 ## Features
 
@@ -92,7 +95,13 @@ In your virtual environment, install ras-commander using pip:
 pip install --upgrade ras-commander
 ```
 If you have dependency issues with pip (especially if you have errors with numpy), try clearing your local pip packages 'C:\Users\your_username\AppData\Roaming\Python\' and then creating a new virtual environment.  
-   
+
+Dependencies can also be manually installed: 
+```
+pip install h5py numpy pandas requests tqdm scipy xarray geopandas matplotlib shapely pathlib rasterstats rtree
+```
+
+
 #### Work in a Local Copy
 
 If you want to make revisions and work actively in your local version of ras-commander, just skip the pip install rascommander step above and clone a fork of the repo to your local machine using Git (ask ChatGPT if you need help).  Most of the notebooks and examples in this repo have a code segment similar to the one below, that works as long as the script is located in a first-level subfolder of the ras-commander repository:
@@ -135,24 +144,27 @@ Using the default 'ras" object, othewise substitute your_ras_project for muli-pr
 ```
 print("\nPlan Files DataFrame:")
 ras.plan_df
-
+```
+```
 print("\nFlow Files DataFrame:")
 ras.flow_df
-
+```
+```
 print("\nUnsteady Flow Files DataFrame:")
 ras.unsteady_df
-
+```
+```
 print("\nGeometry Files DataFrame:")
 ras.geom_df
-
-print("\nHDF Entries DataFrame:")
-ras.get_hdf_entries()
-
+```
+```
 print("\nBoundary Conditions DataFrame:")
 ras.boundaries_df
-
 ```
-
+```
+print("\nHDF Entries DataFrame:")
+ras.get_hdf_entries()
+```
 
 
 
@@ -265,6 +277,11 @@ This is useful for comparing different river systems, running scenario analyses
 
 ```
 ras_commander
+├── ai_tools
+│   ├── [AI Knowledge Bases](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/llm_knowledge_bases) 
+│   └── [Library Assistant](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/library_asssistant)
+├── examples
+│   └── [Examples Notebooks](https://github.com/gpt-cmdr/ras-commander/tree/main/ras_commander)
 ├── ras_commander
 │   ├── __init__.py
 │   ├── _version.py
@@ -277,8 +294,6 @@ ras_commander
 │   ├── RasPrj.py
 │   ├── RasUnsteady.py
 │   ├── RasUtils.py
-│   ├── RasToGo.py
-│   ├── RasGpt.py
 │   ├── HdfBase.py
 │   ├── HdfBndry.py
 │   ├── HdfMesh.py
@@ -292,28 +307,13 @@ ras_commander
 │   ├── HdfFluvialPluvial.py
 │   ├── HdfPlot.py
 │   └── HdfResultsPlot.py
-├── examples
-│   ├── 00_Using_RasExamples.ipynb
-│   ├── 01_project_initialization.ipynb
-│   ├── 02_plan_and_geometry_operations.ipynb
-│   ├── 03_unsteady_flow_operations.ipynb
-│   ├── 04_multiple_project_operations.ipynb
-│   ├── 05_single_plan_execution.ipynb
-│   ├── 06_executing_plan_sets.ipynb
-│   ├── 07_sequential_plan_execution.ipynb
-│   ├── 08_parallel_execution.ipynb
-│   └── 09_plan_parameter_operations.ipynb
-├── tests
-│   └── ... (test files)
 ├── .gitignore
 ├── LICENSE
 ├── README.md
 ├── STYLE_GUIDE.md
 ├── Comprehensive_Library_Guide.md
 ├── pyproject.toml
-├── setup.cfg
 ├── setup.py
-└── requirements.txt
 ```
 
 ### Accessing HEC Examples through RasExamples
@@ -335,6 +335,8 @@ projects = ras_examples.list_projects("Steady Flow")
 extracted_paths = ras_examples.extract_project(["Bald Eagle Creek", "Muncie"])
 ```
 
+The RasExamples class is used to provide an alternative to traditional unit testing, with example notebooks doubling as tests and in-context examples for the end user.  This increases interpretability by LLM's, reducing hallucinations.  
+
 ### RasPrj
 
 The `RasPrj` class is central to managing HEC-RAS projects within the ras-commander library. It handles project initialization, data loading, and provides access to project components.
@@ -459,18 +461,11 @@ The ras-commander library is an ongoing project. Future plans include:
 - [GPT-Commander YouTube Channel](https://www.youtube.com/@GPT_Commander)
 - [ChatGPT Examples for Water Resources Engineers](https://github.com/gpt-cmdr/HEC-Commander/tree/main/ChatGPT%20Examples)
 
-## Contributing
-
-We welcome contributions! Please see our [Contributing Guide](CONTRIBUTING.md) for details on how to submit pull requests, report issues, and suggest improvements.
 
 ## Style Guide
 
 This project follows a specific style guide to maintain consistency across the codebase. Please refer to the [Style Guide](STYLE_GUIDE.md) for details on coding conventions, documentation standards, and best practices.
 
-## License
-
-ras-commander is released under the MIT License. See the license file for details.
-
 ## Acknowledgments
 
 RAS Commander is based on the HEC-Commander project's "Command Line is All You Need" approach, leveraging the HEC-RAS command-line interface for automation. The initial development of this library was presented in the HEC-Commander Tools repository. In a 2024 Australian Water School webinar, Bill demonstrated the derivation of basic HEC-RAS automation functions from plain language instructions. Leveraging the previously developed code and AI tools, the library was created. The primary tools used for this initial development were Anthropic's Claude, GPT-4, Google's Gemini Experimental models, and the Cursor AI Coding IDE.
@@ -488,7 +483,7 @@ Additionally, we would like to acknowledge the following notable contributions a
    Xiaofeng Liu, Ph.D., P.E.,    Associate Professor, Department of Civil and Environmental Engineering
    Institute of Computational and Data Sciences, Penn State University
 
-3. Attribution: The[ffrd\rashdf'](https://github.com/fema-ffrd/rashdf) project by FEMA-FFRD (FEMA Future of Flood Risk Data) was incorporated, revised, adapted and extended in rascommander's RasHDF libaries (where noted). 
+3. Attribution: The [ffrd\rashdf'](https://github.com/fema-ffrd/rashdf) project by FEMA-FFRD (FEMA Future of Flood Risk Data) was incorporated, revised, adapted and extended in rascommander's RasHDF libaries (where noted). 
 
 These acknowledgments recognize the contributions and inspirations that have helped shape RAS Commander, ensuring proper attribution for the ideas and code that have influenced its development.
 
@@ -499,19 +494,25 @@ These acknowledgments recognize the contributions and inspirations that have hel
 ## Official RAS Commander AI-Generated Songs:
 
 [No More Wait and See (Bluegrass)](https://suno.com/song/16889f3e-50f1-4afe-b779-a41738d7617a)  
-
+  
+  
 [No More Wait and See (Cajun Zydeco)](https://suno.com/song/4441c45d-f6cd-47b9-8fbc-1f7b277ee8ed)  
-
+  
 ## Other Resources
 
-Notebook version of RAS-Commander: [RAS-Commander Notebook in the HEC-Commander Tools Repository](https://github.com/gpt-cmdr/HEC-Commander/tree/main/RAS-Commander)
+Notebook version of RAS-Commander: [RAS-Commander Notebook in the HEC-Commander Tools Repository](https://github.com/gpt-cmdr/HEC-Commander/tree/main/RAS-Commander)  
+
 Youtube Tutorials for HEC-Commander Tools and RAS-Commander: [GPT-Commander on YouTube](https://www.youtube.com/@GPT_Commander/videos)
 
+## Contributing
+
+We welcome contributions! Please see our [Contributing Guide](CONTRIBUTING.md) for details on how to submit pull requests, report issues, and suggest improvements.
+
 ## LICENSE
 
 This software is released under the MIT license.
 
 ## Contact
 
-For questions, suggestions, or support, please contact:
+For questions, suggestions, or support, please contact:  
 William Katzenmeyer, P.E., C.F.M. - heccommander@gmail.com

{ras_commander-0.61.0 → ras_commander-0.65.0}/README.md

@@ -3,22 +3,25 @@
 RAS Commander is a Python library for automating HEC-RAS operations, providing a set of tools to interact with HEC-RAS project files, execute simulations, and manage project data. This library was initially conceptualized in the Australian Water School course "AI Tools for Modelling Innovation", and subsequently expanded to cover the basic functionality of the HECRASController COM32 interface using open-source python libraries.  This library uses a Test Driven Development strategy, leveraging the publicly-available HEC-RAS Example projects to create repeatable demonstration examples.  The "Commmander" moniker is inspired by the "Command Line is All You Need" approach to HEC-RAS automation that was first implemented in the HEC-Commander Tools repository. 
 
 ## Repository Author:
-(William Katzenmeyer, P.E., C.F.M.)[https://engineeringwithllms.info]
+[William Katzenmeyer, P.E., C.F.M.](https://engineeringwithllms.info)
+-----  
 
-## Don't Ask Me, Ask GPT! 
+## Don't Ask Me, Ask GPT!
 
 This repository has several methods of interaction with Large Language Models and LLM-Assisted Coding built right in: 
 
-1. **[RAS Commander Library Assistant GPT](https://chatgpt.com/g/g-TZRPR3oAO-ras-commander-library-assistant)**: A specialized GPT model with access to the ras-commander codebase and library, available for answering queries and providing code suggestions.   You can even upload your own plan, unsteady and HDF files to inspect and help determine how to automate your workflows or visualize your results, although this ability is still limited by OpenAI's GPT frameworks and may not be useful for long conversations.
+1. **[Purpose-Built Knowledge Base Summaries](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/llm_knowledge_bases)**: Up-to-date compilations of the documentation and codebase for use with large language models like Claude, ChatGPT, Gemini or Grok. Look in 'ai_tools/assistant_knowledge_bases/' in the repo.  The repo's codebase (without documentation and examples) has been curated to stay within the current ~200k context window limitations of frontier models, and for tasks that do not need an understanding of the underlying code, the Comprehensive Library Guide and any relevant examples from the example folder should be adequate context for leveraging the ras-commander API to complete tasks. 
 
-2. **[Purpose-Built Knowledge Base Summaries](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/assistant_knowledge_bases)**: Up-to-date compilations of the documentation and codebase for use with large language models like Claude or GPT-4. Look in 'ai_tools/assistant_knowledge_bases/' in the repo.  The repo's codebase (without documentation and examples) has been curated to stay within the current ~200k context window limitations of frontier models, and for tasks that do not need an understanding of the underlying code, the Comprehensive Library Guide and any relevant examples from the example folder should be adequate context for leveraging the ras-commander API to complete tasks. 
+2. **[Cursor IDE Integration](https://github.com/gpt-cmdr/ras-commander/blob/main/.cursorrules)**: Custom rules(.cursorrules) for the Cursor IDE to provide context-aware suggestions and documentation.  Just open the repository folder in Cursor to recognize these instructions.  You can create your own folders "/workspace/, "/projects/", or "my_projects/" as these are already in the .gitignore, and place your custom scripts there for your projects.  This will allow easy referencing of the ras-commander documents and individual repo files, the automatic loading of the .cursorrules file.  Alternatvely, download the github repo into your projects folder to easily load documents and use cursor rules files.
 
-3. **[Cursor IDE Integration](https://github.com/gpt-cmdr/ras-commander/blob/main/.cursorrules)**: Custom rules(.cursorrules) for the Cursor IDE to provide context-aware suggestions and documentation.  Just open the repository folder in Cursor to recognize these instructions.  You can create your own folders "/workspace/, "/projects/", or "my_projects/" as these are already in the .gitignore, and place your custom scripts there for your projects.  This will allow easy referencing of the ras-commander documents and individual repo files, the automatic loading of the .cursorrules file.  Alternatvely, download the github repo into your projects folder to easily load documents and use cursor rules files.
+3. **[RAS-Commander Library Assistant](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/library_assistant)**: A full-featured interface for multi-turn conversations, using your own API keys and the ras-commander library for context.  The library assistant allows you to load your own scripts and chat with specific examples and/or function classes in the RAS-Commander library to effectively utilize the library's functions in your workflow.  To reduce hallucinations, a file browser is included which adds full files to the conversation to ensure grounded responses.  A dashboard shows you the total context and estimated cost of each request.  **Now with support for Claude 3.7, OpenAI's o1 and o3-mini, and Deepseek V3 and R1 models using US-based Together.ai**
+
+
+4. **[RAS Commander Library Assistant on ChatGPT](https://chatgpt.com/g/g-TZRPR3oAO-ras-commander-library-assistant)**: A specialized ChatGPT "GPT" with access to the ras-commander codebase and library, available for answering queries and providing code suggestions.   You can even upload your own plan, unsteady and HDF files to inspect and help determine how to automate your workflows or visualize your results.  _NOTE: GPT's are still quite limited by OpenAI's GPT frameworks and may not be useful for long conversations.  Code interpreter cannot run HEC-RAS but can open and view smaller HDF files and projects for demonstration purposes_
 
-5. **[RAS-Commander Library Assistant](https://github.com/gpt-cmdr/ras-commander/blob/main/library_assistant)**: A full-featured interface for multi-turn conversations, using your own API keys and the ras-commander library for context.  The library assistant allows you to load your own scripts and chat with specific examples and/or function classes in the RAS-Commander library to effectively utilize the library's functions in your workflow.  To reduce hallucinations, a file browser is included which adds full files to the conversation to ensure grounded responses.  A dashboard shows you the total context and estimated cost of each request.  **Now with support for OpenAI's o1 and o3-mini, and Deepseek V3 and R1 models using US-based Together.ai**
 
 ## Background
-The ras-commander library emerged from the initial test-bed of AI-driven coding represented by the HEC-Commander tools Python notebooks. These notebooks served as a proof of concept, demonstrating the value proposition of automating HEC-RAS operations. The transition from notebooks to a structured library aims to provide a more robust, maintainable, and extensible solution for water resources engineers.
+The ras-commander library emerged from the initial test-bed of AI-driven coding represented by the [HEC-Commander tools](https://github.com/gpt-cmdr/HEC-Commander) Python notebooks. These notebooks served as a proof of concept, demonstrating the value proposition of automating HEC-RAS operations. In 2024, I taught a series of progressively more complex webinars demonstrating how to use simple prompting, example projects and natural language instruction to effectively code HEC-RAS automation workflows, culminating in a 6 hour course.  The library published for utilization in that course, [awsrastools](https://github.com/gpt-cmdr/awsrastools) served as a foundation of examples which were iteratively extended into the full RAS-Commander library.  Unlike the original notebook by the same name, this library is not focused on parallel execution across multiple machines.  Instead, it is focused on providing a general-purpose python API for interacting with HEC-RAS projects, and building an AI-friendly library that will allow new users to quickly scaffold their own workflows into a python script. Example notebooks are provided, but the intention is to empower engineers, software developers, GIS personnel and data analysts to more easily access and interact with HEC-RAS data in a python environment.  Also, by publishing these examples publicly, with complete working code examples and LLM optimization, future users can readily rewrite they key functions of the library for inclusion in into their own preferred libraries, languages or return formats.
 
 ## Features
 
@@ -60,7 +63,13 @@ In your virtual environment, install ras-commander using pip:
 pip install --upgrade ras-commander
 ```
 If you have dependency issues with pip (especially if you have errors with numpy), try clearing your local pip packages 'C:\Users\your_username\AppData\Roaming\Python\' and then creating a new virtual environment.  
-   
+
+Dependencies can also be manually installed: 
+```
+pip install h5py numpy pandas requests tqdm scipy xarray geopandas matplotlib shapely pathlib rasterstats rtree
+```
+
+
 #### Work in a Local Copy
 
 If you want to make revisions and work actively in your local version of ras-commander, just skip the pip install rascommander step above and clone a fork of the repo to your local machine using Git (ask ChatGPT if you need help).  Most of the notebooks and examples in this repo have a code segment similar to the one below, that works as long as the script is located in a first-level subfolder of the ras-commander repository:
@@ -103,24 +112,27 @@ Using the default 'ras" object, othewise substitute your_ras_project for muli-pr
 ```
 print("\nPlan Files DataFrame:")
 ras.plan_df
-
+```
+```
 print("\nFlow Files DataFrame:")
 ras.flow_df
-
+```
+```
 print("\nUnsteady Flow Files DataFrame:")
 ras.unsteady_df
-
+```
+```
 print("\nGeometry Files DataFrame:")
 ras.geom_df
-
-print("\nHDF Entries DataFrame:")
-ras.get_hdf_entries()
-
+```
+```
 print("\nBoundary Conditions DataFrame:")
 ras.boundaries_df
-
 ```
-
+```
+print("\nHDF Entries DataFrame:")
+ras.get_hdf_entries()
+```
 
 
 
@@ -233,6 +245,11 @@ This is useful for comparing different river systems, running scenario analyses
 
 ```
 ras_commander
+├── ai_tools
+│   ├── [AI Knowledge Bases](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/llm_knowledge_bases) 
+│   └── [Library Assistant](https://github.com/gpt-cmdr/ras-commander/tree/main/ai_tools/library_asssistant)
+├── examples
+│   └── [Examples Notebooks](https://github.com/gpt-cmdr/ras-commander/tree/main/ras_commander)
 ├── ras_commander
 │   ├── __init__.py
 │   ├── _version.py
@@ -245,8 +262,6 @@ ras_commander
 │   ├── RasPrj.py
 │   ├── RasUnsteady.py
 │   ├── RasUtils.py
-│   ├── RasToGo.py
-│   ├── RasGpt.py
 │   ├── HdfBase.py
 │   ├── HdfBndry.py
 │   ├── HdfMesh.py
@@ -260,28 +275,13 @@ ras_commander
 │   ├── HdfFluvialPluvial.py
 │   ├── HdfPlot.py
 │   └── HdfResultsPlot.py
-├── examples
-│   ├── 00_Using_RasExamples.ipynb
-│   ├── 01_project_initialization.ipynb
-│   ├── 02_plan_and_geometry_operations.ipynb
-│   ├── 03_unsteady_flow_operations.ipynb
-│   ├── 04_multiple_project_operations.ipynb
-│   ├── 05_single_plan_execution.ipynb
-│   ├── 06_executing_plan_sets.ipynb
-│   ├── 07_sequential_plan_execution.ipynb
-│   ├── 08_parallel_execution.ipynb
-│   └── 09_plan_parameter_operations.ipynb
-├── tests
-│   └── ... (test files)
 ├── .gitignore
 ├── LICENSE
 ├── README.md
 ├── STYLE_GUIDE.md
 ├── Comprehensive_Library_Guide.md
 ├── pyproject.toml
-├── setup.cfg
 ├── setup.py
-└── requirements.txt
 ```
 
 ### Accessing HEC Examples through RasExamples
@@ -303,6 +303,8 @@ projects = ras_examples.list_projects("Steady Flow")
 extracted_paths = ras_examples.extract_project(["Bald Eagle Creek", "Muncie"])
 ```
 
+The RasExamples class is used to provide an alternative to traditional unit testing, with example notebooks doubling as tests and in-context examples for the end user.  This increases interpretability by LLM's, reducing hallucinations.  
+
 ### RasPrj
 
 The `RasPrj` class is central to managing HEC-RAS projects within the ras-commander library. It handles project initialization, data loading, and provides access to project components.
@@ -427,18 +429,11 @@ The ras-commander library is an ongoing project. Future plans include:
 - [GPT-Commander YouTube Channel](https://www.youtube.com/@GPT_Commander)
 - [ChatGPT Examples for Water Resources Engineers](https://github.com/gpt-cmdr/HEC-Commander/tree/main/ChatGPT%20Examples)
 
-## Contributing
-
-We welcome contributions! Please see our [Contributing Guide](CONTRIBUTING.md) for details on how to submit pull requests, report issues, and suggest improvements.
 
 ## Style Guide
 
 This project follows a specific style guide to maintain consistency across the codebase. Please refer to the [Style Guide](STYLE_GUIDE.md) for details on coding conventions, documentation standards, and best practices.
 
-## License
-
-ras-commander is released under the MIT License. See the license file for details.
-
 ## Acknowledgments
 
 RAS Commander is based on the HEC-Commander project's "Command Line is All You Need" approach, leveraging the HEC-RAS command-line interface for automation. The initial development of this library was presented in the HEC-Commander Tools repository. In a 2024 Australian Water School webinar, Bill demonstrated the derivation of basic HEC-RAS automation functions from plain language instructions. Leveraging the previously developed code and AI tools, the library was created. The primary tools used for this initial development were Anthropic's Claude, GPT-4, Google's Gemini Experimental models, and the Cursor AI Coding IDE.
@@ -456,7 +451,7 @@ Additionally, we would like to acknowledge the following notable contributions a
    Xiaofeng Liu, Ph.D., P.E.,    Associate Professor, Department of Civil and Environmental Engineering
    Institute of Computational and Data Sciences, Penn State University
 
-3. Attribution: The[ffrd\rashdf'](https://github.com/fema-ffrd/rashdf) project by FEMA-FFRD (FEMA Future of Flood Risk Data) was incorporated, revised, adapted and extended in rascommander's RasHDF libaries (where noted). 
+3. Attribution: The [ffrd\rashdf'](https://github.com/fema-ffrd/rashdf) project by FEMA-FFRD (FEMA Future of Flood Risk Data) was incorporated, revised, adapted and extended in rascommander's RasHDF libaries (where noted). 
 
 These acknowledgments recognize the contributions and inspirations that have helped shape RAS Commander, ensuring proper attribution for the ideas and code that have influenced its development.
 
@@ -467,19 +462,25 @@ These acknowledgments recognize the contributions and inspirations that have hel
 ## Official RAS Commander AI-Generated Songs:
 
 [No More Wait and See (Bluegrass)](https://suno.com/song/16889f3e-50f1-4afe-b779-a41738d7617a)  
-
+  
+  
 [No More Wait and See (Cajun Zydeco)](https://suno.com/song/4441c45d-f6cd-47b9-8fbc-1f7b277ee8ed)  
-
+  
 ## Other Resources
 
-Notebook version of RAS-Commander: [RAS-Commander Notebook in the HEC-Commander Tools Repository](https://github.com/gpt-cmdr/HEC-Commander/tree/main/RAS-Commander)
+Notebook version of RAS-Commander: [RAS-Commander Notebook in the HEC-Commander Tools Repository](https://github.com/gpt-cmdr/HEC-Commander/tree/main/RAS-Commander)  
+
 Youtube Tutorials for HEC-Commander Tools and RAS-Commander: [GPT-Commander on YouTube](https://www.youtube.com/@GPT_Commander/videos)
 
+## Contributing
+
+We welcome contributions! Please see our [Contributing Guide](CONTRIBUTING.md) for details on how to submit pull requests, report issues, and suggest improvements.
+
 ## LICENSE
 
 This software is released under the MIT license.
 
 ## Contact
 
-For questions, suggestions, or support, please contact:
-William Katzenmeyer, P.E., C.F.M. - heccommander@gmail.com
+For questions, suggestions, or support, please contact:  
+William Katzenmeyer, P.E., C.F.M. - heccommander@gmail.com

{ras_commander-0.61.0 → ras_commander-0.65.0}/ras_commander/HdfBndry.py

@@ -122,6 +122,12 @@ class HdfBndry:
         -------
         gpd.GeoDataFrame
             A GeoDataFrame containing the breaklines.
+
+        Notes
+        -----
+        - Zero-length breaklines are logged and skipped. 
+        - Single-point breaklines are logged and skipped.
+        - These invalid breaklines should be removed in RASMapper to prevent potential issues.
         """
         try:
             with h5py.File(hdf_path, 'r') as hdf_file:
@@ -129,17 +135,95 @@ class HdfBndry:
                 if breaklines_path not in hdf_file:
                     logger.warning(f"Breaklines path '{breaklines_path}' not found in HDF file.")
                     return gpd.GeoDataFrame()
+
                 bl_line_data = hdf_file[breaklines_path]
-                bl_line_ids = range(bl_line_data["Attributes"][()].shape[0])
-                names = np.vectorize(HdfUtils.convert_ras_string)(
-                    bl_line_data["Attributes"][()]["Name"]
-                )
-                geoms = HdfBase.get_polylines_from_parts(hdf_path, breaklines_path)
+                attributes = bl_line_data["Attributes"][()]
+                
+                # Initialize lists to store valid breakline data
+                valid_ids = []
+                valid_names = []
+                valid_geoms = []
+
+                # Track invalid breaklines for summary
+                zero_length_count = 0
+                single_point_count = 0
+                other_error_count = 0
+
+                # Process each breakline
+                for idx, (pnt_start, pnt_cnt, part_start, part_cnt) in enumerate(bl_line_data["Polyline Info"][()]):
+                    name = HdfUtils.convert_ras_string(attributes["Name"][idx])
+
+                    # Check for zero-length breaklines
+                    if pnt_cnt == 0:
+                        zero_length_count += 1
+                        logger.debug(f"Zero-length breakline found (FID: {idx}, Name: {name})")
+                        continue
+
+                    # Check for single-point breaklines
+                    if pnt_cnt == 1:
+                        single_point_count += 1
+                        logger.debug(f"Single-point breakline found (FID: {idx}, Name: {name})")
+                        continue
+
+                    try:
+                        points = bl_line_data["Polyline Points"][()][pnt_start:pnt_start + pnt_cnt]
+                        
+                        # Additional validation of points array
+                        if len(points) < 2:
+                            single_point_count += 1
+                            logger.debug(f"Invalid point count in breakline (FID: {idx}, Name: {name})")
+                            continue
+
+                        if part_cnt == 1:
+                            geom = LineString(points)
+                        else:
+                            parts = bl_line_data["Polyline Parts"][()][part_start:part_start + part_cnt]
+                            geom = MultiLineString([
+                                points[part_pnt_start:part_pnt_start + part_pnt_cnt]
+                                for part_pnt_start, part_pnt_cnt in parts
+                                if part_pnt_cnt > 1  # Skip single-point parts
+                            ])
+                            # Skip if no valid parts remain
+                            if len(geom.geoms) == 0:
+                                other_error_count += 1
+                                logger.debug(f"No valid parts in multipart breakline (FID: {idx}, Name: {name})")
+                                continue
+
+                        valid_ids.append(idx)
+                        valid_names.append(name)
+                        valid_geoms.append(geom)
+
+                    except Exception as e:
+                        other_error_count += 1
+                        logger.debug(f"Error processing breakline {idx}: {str(e)}")
+                        continue
+
+                # Log summary of invalid breaklines
+                total_invalid = zero_length_count + single_point_count + other_error_count
+                if total_invalid > 0:
+                    logger.info(
+                        f"Breakline processing summary:\n"
+                        f"- Zero-length breaklines: {zero_length_count}\n"
+                        f"- Single-point breaklines: {single_point_count}\n"
+                        f"- Other invalid breaklines: {other_error_count}\n"
+                        f"Consider removing these invalid breaklines using RASMapper."
+                    )
+
+                # Create GeoDataFrame with valid breaklines
+                if not valid_ids:
+                    logger.warning("No valid breaklines found in the HDF file.")
+                    return gpd.GeoDataFrame()
+
                 return gpd.GeoDataFrame(
-                    {"bl_id": bl_line_ids, "Name": names, "geometry": geoms},
+                    {
+                        "bl_id": valid_ids,
+                        "Name": valid_names,
+                        "geometry": valid_geoms
+                    },
                     geometry="geometry",
-                    crs=HdfBase.get_projection(hdf_file),
+                    crs=HdfBase.get_projection(hdf_file)
                 )
+
         except Exception as e:
             logger.error(f"Error reading breaklines: {str(e)}")
             return gpd.GeoDataFrame()

{ras_commander-0.61.0 → ras_commander-0.65.0}/ras_commander/HdfInfiltration.py

@@ -47,7 +47,7 @@ from pathlib import Path
 import pandas as pd
 import geopandas as gpd
 import h5py
-from rasterstats import zonal_stats
+
 from .Decorators import log_call, standardize_input
 
 class HdfInfiltration:
@@ -232,6 +232,12 @@ class HdfInfiltration:
         Returns:
             DataFrame with soil statistics including percentages and areas
         """
+        
+        try:
+            from rasterstats import zonal_stats
+        except ImportError as e:
+            logger.error("Failed to import rasterstats. Please run 'pip install rasterstats' and try again.")
+            raise e
         # Initialize areas dictionary
         mukey_areas = {mukey: 0 for mukey in raster_map.values()}