Step-by-step working of current code for calculating loss and categories

1. Find the control group
   - Identifies the control group by searching for the row that starts with "C " in the 'tableJson' test data.
   - Extracts the values for revenue, conversions, and visitors for this control group.

2. Calculate control metrics
   - Earnings Per Click (EPC): Calculated by dividing revenue by visitors (EPC = revenue ÷ visitors).
   - Conversion Rate (CR): Calculated by dividing conversions by visitors (CR = conversions ÷ visitors).

3. Check each variation (variant)
   - For each variant row:
     - Extracts the values for revenue, conversions, and visitors.
     - Calculates EPC and CR for the variant.
     - Calculates the EPC difference: control EPC - variant EPC.
     - Calculates the Revenue Loss: EPC difference × visitors. This represents the potential money lost if the variant were to run instead of the control.
     - Accumulates totals: Sums up all revenue losses. Compares expected conversions to actual conversions to determine the total conversion loss.

4. Check statistical significance
   - Uses a 95% confidence interval (z = 1.96).
   - Requires a business-defined Minimum Detectable Effect (MDE) of at least 5% impact.
   - A variant's difference is marked as important if it is both statistically significant and exceeds the MDE.
   - A variant is a positive impact if it is better than the control.
   - A variant is a negative impact if it is worse than the control.

5. Decide the test category
   - "Tests to Accept/Implement": Assigned if only positive variants are found.
   - "Tests to Turn Off": Assigned if only negative variants are found.
   - "Undecided Tests": Assigned if there is a mix of positive and negative variants or the results are unclear.

6. Build summary strings
   - Converts numerical data into human-readable strings, such as "Control conversions: 100/200," "Variant A CR: 12.50%," and "Impact: -5.20%."
   - Outputs a summary including total conversions, gained/lost conversions, and monetary impact ($ lost or gained).