Forgot to commit in small portions again

llm_integration/alliance_selection.py

@@ -0,0 +1,15 @@
+import json
+
+
+class AllianceSelectionAssistant:
+    """Assists in alliance selection using LLM and team data."""
+
+    def __init__(self, llm_model):
+        self.llm_model = llm_model
+
+    def generate_pick_list(self, teams_data, strategy):
+        teams_data_str = json.dumps(
+            teams_data
+        )  # Serialize the list of dictionaries to a string
+        prompt = f"Given the following teams data: {teams_data_str} and pick strategy: {pick_strategy}, generate an alliance pick list."
+        return self.query_ollama(prompt)

llm_integration/llm_model.py

@@ -0,0 +1,32 @@
+import json
+import requests
+
+
+class OLLAMAConnector:
+    """
+    Abstract class to interact with a local LLM using Ollama for predictions.
+    """
+
+    def __init__(
+        self, model_name: str, ollama_base_url: str = "http://localhost:11434"
+    ):
+        self.model_name = model_name
+        self.ollama_base_url = ollama_base_url
+
+    def query_ollama(self, prompt: str):
+        """
+        Helper function to query the Ollama API.
+        """
+        url = f"{self.ollama_base_url}/api/generate"
+        data = {
+            "prompt": prompt,
+            "model": self.model_name,
+            "stream": False,  # Set to False to get the full response at once
+        }
+        try:
+            response = requests.post(url, json=data, stream=False)
+            response.raise_for_status()  # Raise HTTPError for bad responses (4xx or 5xx)
+            return response.json()["response"]
+        except requests.exceptions.RequestException as e:
+            print(f"Error querying Ollama: {e}")
+            return None

llm_integration/match_outcome_prediction.py

@@ -0,0 +1,9 @@
+class MatchPredictor:
+    """Predicts match outcomes using LLM."""
+
+    def __init__(self, llm_model):
+        self.llm_model = llm_model
+
+    def predict_outcome(self, blue_alliance_data, red_alliance_data):
+        prompt = f"Given blue alliance data: {blue_alliance_data} and red alliance data: {red_alliance_data}, predict the match outcome."
+        return self.query_ollama(prompt)

llm_integration/team_subjective_rating.py

@@ -0,0 +1,17 @@
+class TeamRatingGenerator:
+    """Generates subjective team ratings using LLM predictions"""
+
+    def __init__(self, llm_model):
+        self.llm_model = llm_model
+
+    def rate_team(
+        self,
+        perfomance_metrics: dict,
+        raw_event_data: dict,
+        isa_data: dict,
+        isa_notes: dict,
+    ) -> str | None:
+        """Rates a team based on available data"""
+        return self.llm_model.query_ollama(
+            prompt=f"The following First Robotics Competition (FRC), data comes from three different sources covering the exact same team and event, please cross reference them to identify possible problems. ```{perfomance_metrics}```, ```{raw_event_data}```, ```{isa_data}``` Do note that while the data source is the same for all three, the presentation of the data doesn't match up perfectly. I also have the following notes about the team in the data: ```{isa_notes}```Once you've done so, please use the data you have collected and referenced to give a comprehensive subjective rating to the team. This is not an interactive conversation, so please give an output that covers everything that you think the user may want in a single message including examples to support the conclusions. THE DATA ONLY CONTAINS ONE TEAM, OUTPUT MUST BE IN HTML FORMAT."
+        )