feat: add statistics

src/app.py

@@ -1,9 +1,13 @@
 """App to display images in a gallery"""
 
+import json
+import pandas as pd
 import numpy as np
 import streamlit as st
 from image_preprocessing import get_image_caption, get_images, resize_image
+import plotly.express as px
 
+from statistics import get_plot_from_most_common_elements
 from data_models.sql_connection import get_db_connection
 from data_models.park_manager import ParkManager
 from data_models.image_manager import ImageManager
@@ -86,6 +90,14 @@ def display_stats() -> None:
     st.write("Number of images: ", image_manager.get_image_count())
     st.write("Number of parks: ", park_manager.get_park_count())
 
+    predictions = openai_manager.get_all_predictions()
+    df = pd.DataFrame(predictions)
+    st.markdown("## Most common elements")
+    st.plotly_chart(get_plot_from_most_common_elements(df, "built_elements", "elements"))
+    st.markdown("## Vegetation detection")
+    st.plotly_chart(get_plot_from_most_common_elements(df, "vegetation_detection", "vegetation"))
+
+
 def main() -> None:
     """Main function to run the app"""
     park = sidebar()

src/data_models/openai_manager.py

@@ -77,6 +77,20 @@ class OpenAIManager:
         except Exception as e:
             raise Exception(f"An error occurred while getting the predictions: {e}")
 
+    def get_all_predictions(self):
+        """
+        Get all predictions from the `openai_predictions` table.
+
+        Returns:
+            list: List of predictions.
+        """
+        query = text("SELECT * FROM openai_predictions")
+        try:
+            result = self.session.execute(query).fetchall()
+            return [row._asdict() for row in result]
+        except Exception as e:
+            raise Exception(f"An error occurred while getting the predictions: {e}")
+
     def close_connection(self):
         """Close the connection."""
         self.session.close()

src/statistics.py

@@ -0,0 +1,38 @@
+import ast
+import itertools
+from collections import Counter
+import pandas as pd
+import plotly.express as px
+from plotly.graph_objs import Figure
+import streamlit as st
+from typing import Any
+
+
+def get_elements(x: str, key: str = "elements") -> Any:
+    try:
+        result = ast.literal_eval(x)
+        if isinstance(result, dict):
+            return result[key]
+        elif isinstance(result, str):
+            return [result]
+        elif isinstance(result, list):
+            return result
+    except:
+        return []
+
+
+def get_most_common_elements(df: pd.DataFrame, column: str,  key: str = "elements") -> list:
+    built_elements = df[column].apply(get_elements, args=(key,))
+    most_common_elements = Counter(itertools.chain.from_iterable(built_elements)).most_common()
+    return most_common_elements
+
+
+def get_plot_from_most_common_elements(df: pd.DataFrame, column: str, key: str = "elements") -> Figure:
+    most_common_elements = get_most_common_elements(df, column, key)
+    most_common_elements = pd.DataFrame(most_common_elements, columns=[column, "count"])
+    return px.bar(
+        most_common_elements,
+        x=column,
+        y="count",
+        labels={"count": "# Objects", column: "Built Elements"},
+    )