When a user starts the Diet module, their initial weight is recorded as the "Measured Weight". 

If the user sets a weight goal, either loss, gain or maintain weight in the app, Measured Weight then becomes the "Estimated Weight". 

The app uses a conversion rate of 7,700 kcal = 1 kg. This means for every 7,700 calories burned beyond intake, a user is expected to lose 1 kg of weight. 

The app adjusts EW daily based on the Daily Budget and Calorie Balance. If a user has a caloric deficit (burns more calories than consumed), CW decreases, and vice versa.

We already talked about the Daily budget and what that entails, however, it’s important for estimated weight, as stated before. 

Daily Budget: 

The budget is calculated based on the RMR, the average number of active calories from the past seven days, and the daily calorie goal based on the user’s weight goal. The Daily Goal is calculated based on your goal weight and date, and it is automatically subtracted from your daily calorie budget.

Another important thing that impacts the Estimated weight is: User Corrector

A "User Corrector" factor adjusts EW based on the accuracy of user input and weigh-ins. It gets activated when the measured weight and estimated weight have difference above 1% for two days in a row. 

Why we need our estimated weight: 

💛Improved Accuracy

💛Enhanced Motivation

💛User-Friendly

How it works: 

Irregularities are detected: The system continuously monitors the user's weight entries. 

Activation of Corrector: Upon detecting a weight spike or drop, the 'corrector' is activated. This tool is designed to analyze the surrounding data points to determine whether the irregular entry is an outlier or part of a new trend.

Recalculation Process:

If the corrector identifies an entry as an outlier, it recalculates the weight data. This recalculation is based on the trend before the irregularity, ensuring continuity in the weight tracking.

The recalculation is also triggered when there is a change in the corrector's settings (e.g., from 1 to 1.03). This ensures that any previously identified irregularities are re-evaluated under the new parameters.

The recalculation process uses the most recent accurate weight entry as a baseline, adjusting subsequent entries to align with the identified trend.

Graphical Representation: All recalculations are reflected in the user's weight graph. Each spike that leads to a recalculation is clearly marked, providing visual feedback to the user.

Continuous Learning: The feature is designed to adapt and improve over time. As it encounters various patterns of irregularities, it fine-tunes its detection and correction algorithms, leading to more accurate adjustments in the future.