Upload 5 files

pages/03 🧑‍💻 Explainable AI.py

@@ -20,6 +20,8 @@ st.image(url, output_format="PNG", width=300)
 
 st.title("Explainable AI")
 
+st.write("For this example, a decision tree model with a max depth of 6 was used.")
+
 df_unclean = pd.read_csv("ifood-data.csv")
 df = df_unclean.dropna()
 df = df[df["Year_Birth"] > 1940]

pages/05 🧑‍💻 Insights.py

@@ -21,3 +21,40 @@ ideal_age = response_by_age.idxmax()
 highest_response_proportion = response_by_age.max()
 
 st.metric(value = ideal_age, label = "Ideal Birth Year")
+
+# Define the number of bins for recency segmentation
+num_bins = 10
+
+# Create bins for recency segmentation
+df['Recency_bins'] = pd.cut(df['Recency'], bins=num_bins)
+
+# Calculate the average response for each recency bin
+average_response = df.groupby('Recency_bins')['Response'].mean()
+
+# Find the recency range with the highest average response
+most_responsive_range = average_response.idxmax()
+
+# Find the average rececny within the most responsive range
+most_responsive_recency_avg = df.loc[df['Recency_bins'] == most_responsive_range, 'Recency'].mean()
+
+st.metric(value = str(round(most_responsive_recency_avg, 2)), label = "Days since last visit")
+
+df['NumWebVisitsMonth_bins'] = pd.cut(df['NumWebVisitsMonth'], bins=num_bins)
+
+average_response = df.groupby('NumWebVisitsMonth')['Response'].mean()
+
+most_responsive_range = average_response.idxmax()
+
+most_responsive_NumWebVisitsMonth_avg = df.loc[df['NumWebVisitsMonth'] == most_responsive_range, 'NumWebVisitsMonth'].mean()
+
+st.metric(value = str(round(most_responsive_NumWebVisitsMonth_avg, 2)), label = "Web visits per month")
+
+df['MntWines_bins'] = pd.cut(df['MntWines'], bins=num_bins)
+
+average_response = df.groupby('MntWines')['Response'].mean()
+
+most_responsive_range = average_response.idxmax()
+
+most_responsive_MntWines_avg = df.loc[df['MntWines'] == most_responsive_range, 'MntWines'].mean()
+
+st.metric(value = "$"+ str(round(most_responsive_MntWines_avg, 2)), label = "Amount spent on wine")