Scenario Comparison

Estimated Time for Completion: 15 minutes; Difficulty Level: Advanced

Taipy offers a way to compare data nodes across scenarios by including a function directly in the configuration of the scenario. This is particularly useful to visualize the differences between scenarios and make decision upon it.

Download the code

Comparing scenarios

Step 1: Declare data nodes on which the comparison functions are applied.

In this example, we want to apply a comparison to the 'revenues' Data Node. It is indicated in the comparators parameter of the configure_scenario.

# Scenario configuration
scenario_cfg = Config.configure_scenario(
    id="pricing_strategy",
    task_configs=[predict_sales_task_cfg, calculate_revenue_task_cfg],
    comparators={revenue_output_cfg.id: compare_revenue}
)

Step 2: Implement the comparison function (compare_function()) used above.

revenues is the list of the Output (revenues) Data Nodes from all scenarios passed in the comparator. We iterate through it to compare scenarios.

def compare_revenue(*revenues):
    scenario_names = [f"Scenario {i}" for i in range(len(revenues))]
    comparisons = {"Scenarios": scenario_names,
                   "Revenues": list(revenues)}
    return pd.DataFrame(comparisons)

Now, the compare_scenarios can be used within Taipy.

if __name__ == "__main__":
    orchestrator = tp.Orchestrator()
    orchestrator.run()

    # Create scenarios with different pricing strategies
    scenario_1 = tp.create_scenario(scenario_cfg)
    scenario_2 = tp.create_scenario(scenario_cfg)

    scenario_1.price_input.write(120)  # Higher price scenario
    scenario_2.price_input.write(80)   # Lower price scenario

    scenario_1.submit()
    scenario_2.submit()

    # Compare the scenarios
    comparisons = tp.compare_scenarios(scenario_1, scenario_2)
    print(comparisons)

Results:

...
{'revenue_output': {'compare_revenue':     Scenarios  Revenues
0  Scenario 0    9120.0
1  Scenario 1   11200.0}}

Compare Scenario with a GUI

Taipy can then generate a GUI that compares scenarios easily. Our comparison will be shown through a chart and a table in this case. This application can be as interactive and complete as possible.

from taipy.gui import Gui
import taipy.gui.builder as tgb

with tgb.Page() as compare_page:
    tgb.text("# Compare Scenarios", mode="md")

    tgb.chart("{comparisons_revenue}", type="bar", x="Scenarios", y="Revenues")
    tgb.table("{comparisons_revenue}")

Gui(compare_page).run()

Entire code

Here is the entire code to recreate the example. A GUI has been created to show the results.

import numpy as np
import pandas as pd

import taipy as tp
import taipy.gui.builder as tgb
from taipy import Config
from taipy.gui import Gui


# Simulation function to predict sales based on pricing
def predict_sales(price):
    """
    Simulate sales volume based on price.
    Elasticity affects how demand reacts to price changes.
    """
    base_demand = 100
    elasticity = -1.5

    demand = base_demand * (price / 100) ** elasticity
    return max(0, np.round(demand))  # Ensure non-negative sales volume

# Function to calculate revenue from sales volume and price
def calculate_revenue(price, sales_volume):
    return price * sales_volume

# Comparator function to compare revenue outputs
def compare_revenue(*revenues):
    scenario_names = [f"Scenario {i}" for i in range(len(revenues))]
    comparisons = {"Scenarios": scenario_names,
                   "Revenues": list(revenues)}
    return pd.DataFrame(comparisons)

if __name__=="__main__":
    # Data Node configuration
    price_input_cfg = Config.configure_data_node("price_input", default_data=100)
    sales_output_cfg = Config.configure_data_node("sales_output")
    revenue_output_cfg = Config.configure_data_node("revenue_output")

    # Task configurations
    predict_sales_task_cfg = Config.configure_task(
        id="predict_sales",
        function=predict_sales,
        input=price_input_cfg,
        output=sales_output_cfg
    )

    calculate_revenue_task_cfg = Config.configure_task(
        id="calculate_revenue",
        function=calculate_revenue,
        input=[price_input_cfg, sales_output_cfg],
        output=revenue_output_cfg
    )

    # Scenario configuration
    scenario_cfg = Config.configure_scenario(
        id="pricing_strategy",
        task_configs=[predict_sales_task_cfg, calculate_revenue_task_cfg],
        comparators={revenue_output_cfg.id: compare_revenue}
    )

    orchestrator = tp.Orchestrator()
    orchestrator.run()

    # Create scenarios with different pricing strategies
    scenario_1 = tp.create_scenario(scenario_cfg)
    scenario_2 = tp.create_scenario(scenario_cfg)

    scenario_1.price_input.write(120)  # Higher price scenario
    scenario_2.price_input.write(80)   # Lower price scenario

    scenario_1.submit()
    scenario_2.submit()

    # Compare the scenarios
    comparisons = tp.compare_scenarios(scenario_1, scenario_2)
    print(comparisons)

    comparisons_revenue = comparisons["revenue_output"]['compare_revenue']

    with tgb.Page() as compare_page:
        tgb.text("# Compare Scenarios", mode="md")

        tgb.chart("{comparisons_revenue}", type="bar", x="Scenarios", y="Revenues")
        tgb.table("{comparisons_revenue}")

    Gui(compare_page).run()