Merge pull request drudilorenzo#9 from batterylake/chowington-large-s…

…cale-hide-and-seek-dev-merge

Chowington large scale hide and seek dev merge

.gitignore

@@ -10,6 +10,9 @@ environment/frontend_server/storage/*
 !environment/frontend_server/storage/July1_the_ville_isabella_maria_klaus-step-3-*/
 !environment/frontend_server/storage/test-22/
 !environment/frontend_server/storage/test-23/
+environment/frontend_server/storage/*/environment/*
+!environment/frontend_server/storage/*/environment/0.json
+environment/frontend_server/storage/*/movement/*
 
 bin/
 include/

environment/frontend_server/storage/base_the_ville_isabella_maria_klaus/environment/0.json

@@ -15,12 +15,3 @@
     "y": 57
   }
 }
-
-
-
-
-
-
-
-
-

environment/frontend_server/storage/base_the_ville_isabella_maria_klaus/plugins/hide-and-seek/prompt_template/hide-and-seek.txt

@@ -0,0 +1,32 @@
+hide-and-seek.txt
+
+Variables: 
+!<INPUT 0>! -- persona 1 learned
+!<INPUT 1>! -- persona 2 learned
+!<INPUT 2>! -- current players status
+!<INPUT 3>! -- current conversation
+!<INPUT 4>! -- persona 1 name
+!<INPUT 5>! -- persona 2 name
+<commentblockmarker>###</commentblockmarker>
+Context for the task: 
+
+The following 2 people are playing a game of hide-and-seek.
+!<INPUT 0>!
+!<INPUT 1>!
+
+Current players status:
+!<INPUT 2>!
+
+Current conversation:
+!<INPUT 3>!
+
+---
+Task: Given the above, decide whether a game of hide-and-seek just ended, and whether each person won. If a game is still going, no one has won yet.
+Output format: Output a json of the following format: 
+{
+    "Did a game of hide-and-seek just end?": "<json Boolean>"
+    "Did !<INPUT 4>! win?": "<json Boolean>",
+    "Did !<INPUT 5>! win?": "<json Boolean>",
+}
+
+Output:

nlp/auto_keyword_count.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+#Script is inted to be used to automate word count processing
+# Define an array to hold the file names
+files=()
+# Define the directory path
+dir="convo-analysis/JSON/"
+
+# Loop through each file in the specified directory
+for file in "$dir"*; do
+    # Check if it's a regular file (not a directory or link, etc.)
+    if [ -f "$file" ]; then
+        # Add the file to the array, extracting just the filename if needed
+        files+=("$(basename "$file")")
+    fi
+done
+
+
+# Loop through the array of files
+for file in "${files[@]}"; do
+    python3 analysis_convo.py "$file"
+
+done
+
+

nlp/auto_raw_text.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+#Script is inted to be used to automate raw text processing
+# Define an array to hold the file names
+files=()
+# Define the directory path
+dir="convo-analysis/JSON/"
+
+# Loop through each file in the specified directory
+for file in "$dir"*; do
+    # Check if it's a regular file (not a directory or link, etc.)
+    if [ -f "$file" ]; then
+        # Add the file to the array, extracting just the filename if needed
+        files+=("$(basename "$file")")
+    fi
+done
+
+
+# Loop through the array of files
+for file in "${files[@]}"; do
+    python3 raw_text_scrape.py "$file"
+
+done
+
+

nlp/auto_topic_modeling.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+#Script is inted to be used to automate topic modeling processing
+# Define an array to hold the file names
+files=()
+# Define the directory path
+dir="convo-analysis/raw-text-convo/"
+
+# Loop through each file in the specified directory
+for file in "$dir"*; do
+    # Check if it's a regular file (not a directory or link, etc.)
+    if [ -f "$file" ]; then
+        # Add the file to the array, extracting just the filename if needed
+        files+=("$(basename "$file")")
+    fi
+done
+
+
+# Loop through the array of files
+for file in "${files[@]}"; do
+    python3 topic_modeling.py "$file"
+
+done
+
+

nlp/key_words.py

@@ -0,0 +1,73 @@
+# This script is intended to analyze agent conversations
+
+# importing modules
+import json
+import sys
+
+
+# Function to analyze conversations and search for specific keywords
+def analyze_convo(json_file, json_data, keywords):
+    non_rep_convos = set()
+    # Iterate through each entry in the JSON data
+    for entry in json_data:
+        # Check each persona's conversation if available
+        for persona, details in entry["persona"].items():
+            if "chat" in details and details["chat"]:
+                for conversation in details["chat"]:
+                    # Check if any of the keywords are in the conversation
+                    for keyword in keywords:
+                        if conversation[1] in non_rep_convos:
+                            continue
+                        if keyword in conversation[1]:
+                            non_rep_convos.add(
+                                f"Keyword '{keyword}' found in conversation of {conversation[0]}: {conversation}"
+                            )
+                            keywords[keyword] += 1
+
+    with open(f"convo-analysis/key-words/{json_file}.txt", "w") as newfile:
+        newfile.write("Keywords in Agent Conversations:\n")
+        for element in non_rep_convos:
+            newfile.write(f"{element}\n")
+        newfile.write("\n")
+        newfile.write(f"Keyword Count:\n")
+        for key in keywords.keys():
+            newfile.write(f"{key}: {keywords[key]}\n")
+
+
+def main():
+    try:
+        # Load the JSON data
+        json_file = sys.argv[1]
+        step_file = f"convo-analysis/JSON/{json_file}"
+
+        with open(step_file, "r") as file:
+            json_data = json.load(file)
+
+        # Define the keywords to search for and keeping a number count
+        keywords = {
+            "hiding": 0,
+            "hide": 0,
+            "trick": 0,
+            "tricky": 0,
+            "search": 0,
+            "find": 0,
+            "found you": 0,
+            "found": 0,
+        }
+
+        # Analyze the conversations
+        analyze_convo(json_file, json_data, keywords)
+
+    except IndexError:
+        print("Error: No file name provided.")
+    except FileNotFoundError:
+        print(f"Error: File '{json_file}' not found.")
+    except json.JSONDecodeError:
+        print("Error: Invalid JSON file.")
+        print(json_file)
+    except Exception as e:
+        print(f"An unexpected error occurred: {e}")
+
+
+if __name__ == "__main__":
+    main()

nlp/lexical_diversity.py

@@ -0,0 +1,18 @@
+from nltk.tokenize import word_tokenize
+
+
+def lexical_diversity(text):
+    return len(set(text)) / len(text)
+
+
+file = "test.txt"
+
+with open(file, "r") as f:
+    # Tokenize and remove punctuation
+    content = f.read()
+    tokens = word_tokenize(content)
+    tokens = [word.lower() for word in tokens if word.isalpha()]
+
+    # Calculate lexical diversity
+    lexical_diversity_score = lexical_diversity(tokens)
+    print(f"Lexical Diversity: {lexical_diversity_score}")

nlp/openai_convo_summary.py

@@ -0,0 +1,57 @@
+# This script is intended to be used to have openai summarize our agent
+# conversations and determine for use the most relevent topics
+
+##import libraries
+import sys
+from openai import OpenAI
+from reverie.backend_server.utils import *
+
+client = OpenAI(api_key="Your API Key Here")
+
+
+# Function declaration
+def file_chuncking(file_path):
+    chunk_size = 1024 * 1024
+    chunks = []
+
+    # Open the large file in read mode
+    with open(file_path, "rb") as file:
+        chunk = file.read(chunk_size)  # Read the first chunk
+
+        while chunk:
+            chunks.append(chunk)  # Add the chunk to the list
+            chunk = file.read(chunk_size)  # Read the next chunk
+
+    return chunks
+
+
+def main():
+    raw_file = sys.argv[1]
+
+    pathtofile = f"convo-analysis/raw-text-convo/{raw_file}"
+
+    file_chunks = file_chuncking(pathtofile)
+    print(file_chunks[0])
+
+    response = client.chat.completions.create(
+        model="gpt-4",
+        messages=[
+            {
+                "role": "system",
+                "content": """ 
+            You are an expert natural language processing expert. You will be provided with the text data that is the conversation 
+            between two generative agents. You will conduct a topic modeling analysis on the agent conversations. Then you will 
+            report the analysis and a concise manner with detailed information.""",
+            },
+            {"role": "user", "content": f"{file_chunks[0]}"},
+        ],
+        temperature=0.7,
+        max_tokens=4096,
+        top_p=1,
+    )
+    print()
+    print(response)
+
+
+if __name__ == "__main__":
+    main()

nlp/raw_text_scrape.py

@@ -0,0 +1,49 @@
+# This script is intended to scape agent conversations only
+
+# importing modules
+import json
+import sys
+
+
+# Function to analyze conversations and search for specific keywords
+def raw_text_scrape(json_data, json_file):
+    raw_data = []
+    # Iterate through each entry in the JSON data
+    for entry in json_data:
+        # Check each persona's conversation if available
+        for persona, details in entry["persona"].items():
+            if "chat" in details and details["chat"]:
+                for conversation in details["chat"]:
+                    raw_data.append(conversation[1])
+                    # print(f"{conversation[1]}\n")
+
+    with open(f"convo-analysis/raw-text-convo/{json_file[:-5]}.txt", "w") as newfile:
+        for i in range(len(raw_data)):
+            newfile.write(f"{raw_data[i]}\n")
+
+
+def main():
+    try:
+        # Load the JSON data
+        json_file = sys.argv[1]
+        step_file = f"convo-analysis/JSON/{json_file}"
+
+        with open(step_file, "r") as file:
+            json_data = json.load(file)
+
+        # Analyze the conversations
+        raw_text_scrape(json_data, json_file)
+
+    except IndexError:
+        print("Error: No file name provided.")
+    except FileNotFoundError:
+        print(f"Error: File '{json_file}' not found.")
+    except json.JSONDecodeError:
+        print("Error: Invalid JSON file.")
+        print(json_file)
+    except Exception as e:
+        print(f"An unexpected error occurred: {e}")
+
+
+if __name__ == "__main__":
+    main()