CivilEngineerUK

"""
# OpenAI Chat Completions Serialization Example

## Description
This code snippet demonstrates how to serialize the response from the OpenAI Chat Completions API into a JSON format which is useful for sending using REST.
It includes a helper function, `serialize_completion`, that takes the completion response object and converts it into a JSON-compatible dictionary.

## Usage
To use this code, you will need to have a .env file in the same directory as this script with your openai key:
OPENAI_API_KEY = "sk-your_key_here"

CivilEngineerUK

"""
Title: Directory Loader using Unstructured API with Error Handling and LangChain Document Output

Description: A class for loading documents from a directory using the LangChain format.
It supports recursive loading and conversion of documents to the LangChain Document class using the UnstructuredClient API.
The class includes built-in error handling and outputs documents in the same style as the LangChain Document class.

To use, get a free unstructured API key here: https://unstructured.io/api-key

Author: @CivilEngineerUK

CivilEngineerUK

from pydantic import BaseModel, Field
from typing import Optional, List
from instructor import OpenAISchema
import json


class Variable(BaseModel):
    name: str = Field(..., description="Name of the variable in markdown math")
    description: str = Field(..., description="Description of the variable")
    unit: Optional[str] = Field(None, description="Unit of the variable")

CivilEngineerUK

"""
Title: Asynchronous OpenAI API Wrapper with Error Handling

Description: A Python script to interact with the OpenAI API asynchronously, with built-in error handling and retry logic.
This script is designed to handle multiple prompts with the ability to define custom functions for more complex queries.
Created from guidance at: https://platform.openai.com/docs/guides/gpt/function-calling

Updated: rev 1 - 16/11/2023 for new end point and gpt-4-1106-preview [openai==1.2.4]
               - includes structural engineering beam formula schema and use of Pydantic and instructor

CivilEngineerUK

#' R_interact_Excel
#' 
#' Allows a spreadsheet to be modified programatically and the 
#'   output from any number of dependent cells returned as a \code{tibble}. 
#'   Uses the \code{RDCOMClient}, \code{dplyr} and \code{stringr} packages
#'   
#' @param file_location file location of the spreadsheet to modify. The
#'   spreadsheet should be \code{.xls} format
#' @param input_values a vector of values corresponding to the cells to be modified
#' @param input_cells a character vector of the input cells