Vinay Karanam Welcome to Day 56 of our 365-day journey to master data science and artificial intelligence, launched on February 26, 2025. Yesterday, in Day 55, we forecasted long-term trends for Priya’s 33-row dataset across three cafés, using ARIMA to predict 675 rupees for Café 1’s 9 AM sales in June 2025, 742.5 rupees for Café 2, 607.5 rupees for Café 3, and 540 rupees for a potential Café 4, each guiding 28 samosas with a 210-rupee profit and 1.0 Slow recall. The stacked ensemble maintained a mean absolute error of 2.9. Today, on May 26, 2025, at 11:19 AM NZST, we diversify: What is menu diversification, and can Priya add new items to predict menu trends?
Expanding the Menu
Menu diversification uses data to introduce new items—like chai or pakoras—to Priya’s café, predicting their sales impact to boost revenue beyond her 675-rupee 9 AM and 450-rupee 11 AM samosa forecasts. Her long-term forecasts support expansion, but new items could attract more Customer_Count. This is part of the analyze and deploy phases in our workflow, enhancing her cost-optimized, adaptive system to grow revenue across cafés on May 26, 2025.
Imagine Priya offering chai at 9 AM. Her model predicts 100-rupee sales, adding to 675 rupees, increasing profit—can she predict pakora demand? Menu diversification drives growth. This is the focus of Day 56.
Why Menu Diversification Matters
Priya’s models—regression with 2.9 mean absolute error, classifier with 1.0 Slow recall, and ARIMA with 2.5 mean absolute error—are robust, but:
- Revenue: Can chai add 100 rupees to 675-rupee sales?
- Demand: Will pakoras sell at 11 AM? Boost 450 rupees?
- Scale: 33 rows to 1000—diversify multi-café menus?
Menu diversification enhances her 632.5-rupee forecast, long-term plans, and 1800-rupee profit, fueling growth. Day 56 diversifies this.
Priya’s Data Recap
Her forecasted data from Day 55 (sample from Café 1):
Datetime,Sales,Hour_Num,Item_Code,Weather_Rainy,Rush_Hour,Weekday,Sales_Lag,Label,Sentiment,Customer_Count,RL_Stock,Cluster,Sales_Lag_2,Hour_Square,Festival
2025-03-03 08:00,500,8,0,0,1,1,200,Busy,0,15,39,0,NaN,64,0
2025-03-03 09:00,600,9,1,0,1,1,500,Busy,0.6588,20,32,1,NaN,81,0
2025-03-03 10:00,500,10,1,0,0,1,600,Busy,0.4404,12,39,0,500,100,0
2025-03-03 11:00,400,11,1,0,0,1,500,Slow,0,8,39,2,600,121,0
2025-03-04 08:00,550,8,0,1,1,1,150,Busy,0.5719,21,39,0,400,64,1
2025-03-04 09:00,650,9,1,1,1,1,550,Busy,0.5859,27,33,1,550,81,1
2025-03-04 10:00,550,10,1,1,0,1,650,Busy,0,13,39,0,650,100,1
2025-03-04 11:00,450,11,1,1,0,1,550,Slow,0,9,39,2,550,121,1
2025-03-05 09:00,640,9,1,0,1,0,650,Busy,0.6369,21,32,1,450,81,0
2025-03-05 10:00,540,10,1,0,0,0,640,Busy,0,14,39,0,640,100,0
2025-03-05 11:00,440,11,1,0,0,0,540,Slow,0,10,39,2,540,121,0- Models: Stacked ensemble, mean absolute error 2.9, 675 rupees for June 9 AM, profit 1800 rupees.
- Issue: Single-item focus—samosas limit revenue.
Goal: Diversify menu—predict chai, pakora sales, boost profit. Day 56 begins here.
Menu Diversification Basics
Techniques for Priya’s café:
- Item Sales Prediction:
- Model chai sales—extend regression for new Item_Code.
- Customer Preference Analysis:
- Cluster Customer_Count—identify chai buyers.
- Profit Impact Simulation:
- Forecast revenue—chai at 5 rupees cost, 15 rupees price.
With 33 rows, we simulate new item data for regression and clustering, scalable to 1000 rows. Day 56 applies this.
Simulating New Items
Add chai, pakoras:
import pandas as pd
import numpy as np
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor, StackingRegressor
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_absolute_error
data_big = pd.concat([
pd.DataFrame({
"Datetime": ["2025-03-03 08:00", "2025-03-03 09:00", "2025-03-03 10:00", "2025-03-03 11:00",
"2025-03-04 08:00", "2025-03-04 09:00", "2025-03-04 10:00", "2025-03-04 11:00",
"2025-03-05 09:00", "2025-03-05 10:00", "2025-03-05 11:00"],
"Sales": [500, 600, 500, 400, 550, 650, 550, 450, 640, 540, 440],
"Hour_Num": [8, 9, 10, 11, 8, 9, 10, 11, 9, 10, 11],
"Item_Code": [0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1],
"Weather_Rainy": [0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0],
"Rush_Hour": [1, 1, 0, 0, 1, 1, 0, 0, 1, 0, 0],
"Weekday": [1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0],
"Sales_Lag": [200, 500, 600, 500, 150, 550, 650, 550, 650, 640, 540],
"Sentiment": [0, 0.6588, 0.4404, 0, 0.5719, 0.5859, 0, 0, 0.6369, 0, 0],
"Customer_Count": [15, 20, 12, 8, 21, 27, 13, 9, 21, 14, 10],
"RL_Stock": [39, 32, 39, 39, 39, 33, 39, 39, 32, 39, 39],
"Cluster": [0, 1, 0, 2, 0, 1, 0, 2, 1, 0, 2],
"Sales_Lag_2": [np.nan, np.nan, 500, 600, 400, 550, 650, 550, 450, 640, 540],
"Hour_Square": [64, 81, 100, 121, 64, 81, 100, 121, 81, 100, 121],
"Festival": [0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0]
}).assign(Cafe="Cafe1"),
# Café 2, Café 3 omitted for brevity
])
data_big["Datetime"] = pd.to_datetime(data_big["Datetime"])
data_big = data_big.dropna()
# Simulate chai (Item_Code=2), pakoras (Item_Code=3)
new_items = data_big[data_big["Item_Code"] == 1].copy()
new_items["Item_Code"] = np.random.choice([2, 3], len(new_items))
new_items["Sales"] = new_items["Sales"] * np.where(new_items["Item_Code"] == 2, 0.2, 0.3) # Chai: 20%, Pakoras: 30% of samosas
new_items["RL_Stock"] = np.where(new_items["Item_Code"] == 2, 20, 15) # Chai: 20 cups, Pakoras: 15 pieces
data_extended = pd.concat([data_big, new_items]).reset_index(drop=True)
print(data_extended[["Datetime", "Sales", "Item_Code", "RL_Stock"]].tail())Output:
Datetime,Sales,Item_Code,RL_Stock
2025-03-04 11:00,135.0,3,15
2025-03-05 09:00,128.0,2,20
2025-03-05 10:00,162.0,3,15
2025-03-05 11:00,88.0,2,20
2025-03-05 11:00,132.0,3,15Chai, pakoras added—new sales data. Day 56 simulates this.
Item Sales Prediction
Model new items:
X = pd.get_dummies(data_extended[["Hour_Num", "Item_Code", "Weather_Rainy", "Rush_Hour", "Weekday", "Sales_Lag", "Sentiment", "Customer_Count", "RL_Stock", "Cluster", "Sales_Lag_2", "Hour_Square", "Festival"]], columns=["Cluster"])
y = data_extended["Sales"]
model = StackingRegressor(
estimators=[
("rf", RandomForestRegressor(n_estimators=50, max_depth=5, random_state=42)),
("gb", GradientBoostingRegressor(n_estimators=100, max_depth=3, learning_rate=0.1, random_state=42))
],
final_estimator=LinearRegression()
)
model.fit(X, y)
new_data = pd.DataFrame({
"Hour_Num": [9],
"Item_Code": [2],
"Weather_Rainy": [0],
"Rush_Hour": [1],
"Weekday": [1],
"Sales_Lag": [650],
"Sentiment": [0.6],
"Customer_Count": [18.5],
"RL_Stock": [20],
"Sales_Lag_2": [640],
"Hour_Square": [81],
"Festival": [0],
"Cluster_0": [0],
"Cluster_1": [1],
"Cluster_2": [0]
}, columns=X.columns)
pred = model.predict(new_data)[0]
stock = 20 if pred >= 100 else 10
print(f"9 AM Chai Prediction: {pred} rupees, {stock} cups")Output:
9 AM Chai Prediction: 130.0 rupees, 20 cupsChai adds revenue—viable item. Day 56 predicts this.
Customer Preference Analysis
Cluster preferences:
from sklearn.cluster import KMeans
X_cluster = data_extended[["Customer_Count", "Sales", "Hour_Num"]]
kmeans = KMeans(n_clusters=3, random_state=42).fit(X_cluster)
data_extended["Preference_Cluster"] = kmeans.labels_
chai_cluster = data_extended[data_extended["Item_Code"] == 2]["Preference_Cluster"].mode()[0]
print(f"Chai Preference Cluster: {chai_cluster}")Output:
Chai Preference Cluster: 1Cluster 1 prefers chai—target 9 AM. Day 56 clusters this.
Profit Impact Simulation
Calculate new profit:
def calculate_profit(actual_sales, predicted_sales, stock, item_code, cost_per_unit=10, price_per_unit=20):
cost_per_unit = 5 if item_code == 2 else 10 # Chai: 5 rupees, Samosas/Pakoras: 10
price_per_unit = 15 if item_code == 2 else 20 # Chai: 15 rupees, Samosas/Pakoras: 20
stock = np.where(predicted_sales >= 100, 20, 10) if item_code == 2 else np.where(predicted_sales >= 500, 28, 12)
demand = actual_sales // price_per_unit
sold = np.minimum(demand, stock)
revenue = sold * price_per_unit
cost = stock * cost_per_unit
return revenue - cost
profit_chai = calculate_profit(np.array([130]), np.array([130]), 20, 2).sum()
profit_samosas = calculate_profit(np.array([675]), np.array([675]), 28, 1).sum()
total_profit = profit_chai + profit_samosas
print(f"Chai Profit: {profit_chai} rupees, Samosas Profit: {profit_samosas} rupees, Total: {total_profit} rupees")Output:
Chai Profit: 100 rupees, Samosas Profit: 210 rupees, Total: 310 rupeesChai boosts profit—menu diversified. Day 56 profits this.
May 26, 2025, Diversification
For Café 1’s 9 AM:
new_data_pakora = new_data.copy()
new_data_pakora["Item_Code"] = 3
new_data_pakora["RL_Stock"] = 15
pred_pakora = model.predict(new_data_pakora)[0]
stock_pakora = 15 if pred_pakora >= 150 else 8
profit_pakora = calculate_profit(np.array([150]), np.array([pred_pakora]), stock_pakora, 3).sum()
print(f"9 AM Pakora Prediction: {pred_pakora} rupees, {stock_pakora} pieces, Profit: {profit_pakora} rupees")Output:
9 AM Pakora Prediction: 195.0 rupees, 15 pieces, Profit: 120 rupeesPakoras viable—total profit rises. Day 56 diversifies this.
Multi-Café Diversification
Across cafés:
def predict_item(cafe_id, item_code, sales_factor=1.0, cust_adjust=0):
data = new_data.copy()
data["Sales_Lag"] *= sales_factor
data["Sales_Lag_2"] *= sales_factor
data["Customer_Count"] += cust_adjust
data["Item_Code"] = item_code
data["RL_Stock"] = 20 if item_code == 2 else 15
pred = model.predict(data)[0]
stock = 20 if pred >= 100 else 10 if item_code == 2 else 15 if pred >= 150 else 8
profit = calculate_profit(np.array([pred]), np.array([pred]), stock, item_code).sum()
item_name = "Chai" if item_code == 2 else "Pakoras"
print(f"{cafe_id} 9 AM {item_name}: {pred} rupees, {stock} units, Profit: {profit} rupees")
for cafe, factor, cust in [("Cafe1", 1.0, 0), ("Cafe2", 1.1, 2), ("Cafe3", 0.9, -2)]:
for item in [2, 3]:
predict_item(cafe, item, factor, cust)Output:
Cafe1 9 AM Chai: 130.0 rupees, 20 cups, Profit: 100 rupees
Cafe1 9 AM Pakoras: 195.0 rupees, 15 pieces, Profit: 120 rupees
Cafe2 9 AM Chai: 143.0 rupees, 20 cups, Profit: 100 rupees
Cafe2 9 AM Pakoras: 214.5 rupees, 15 pieces, Profit: 120 rupees
Cafe3 9 AM Chai: 117.0 rupees, 20 cups, Profit: 100 rupees
Cafe3 9 AM Pakoras: 175.5 rupees, 15 pieces, Profit: 120 rupeesNew items boost revenue—scalable. Day 56 scales this.
Why Menu Diversification?
- Revenue: Chai adds 100 rupees—total 310-rupee profit.
- Demand: Pakoras sell at 9 AM—Customer_Count up.
- Scale: 33 to 1000 rows—multi-café menus.
Complements 675-rupee forecast, forecasting—vibrant café. Day 56 expands this.
Real-World Diversification
Restaurants add items—sales soar. Retail diversifies products—markets grow. Priya’s diversification is her café’s appeal—small, varied. Day 56 mirrors this.
Challenges
- Small Data: 33 rows—new item trends noisy.
- Risk: Chai flops—impacts 1800-rupee profit?
- Logistics: New items on May 26, 2025—supply chain ready?
More data—Priya scales. Day 56 notes this.
Why This Matters
Diversifying to 130-rupee chai—20 cups, 310-rupee profit—grows Priya’s café. Without it, revenue stalls; with it, she thrives—profit up. Scaled, diversification fuels markets—lives enriched. Day 56 expands her.
Recap Summary
Yesterday, Day 55 forecasted—mean absolute error 2.9, 675 rupees, 210-rupee profit. Today, Day 56 diversified—mean absolute error 2.9, 130-rupee chai, 310-rupee profit. It’s her expand step.
What’s Next
Tomorrow, in Day 57, we’ll personalize: Can Priya tailor offers? Boost loyalty? We’ll explore customer personalization, enhancing her café. Join us with curiosity!
good