GR:eat Prediction Algorithm Tester

Meal Presets

Current Meal Macros

Carbs (g)

Protein (g)

Fat (g)

Fiber (g)

Activity Level

Meal Hour (0-23)

Current CGM Reading (optional, for calibration)

Meal Stacking (Active Meals)

▼

Add previous meals still digesting (within 4h window).

Carbs

Protein

Fat

Fiber

Min Ago

User Profile

▼

Age

Fasting Glucose

Weight (kg)

Height (cm)

HbA1c

Diabetes Status

Circadian Glucose Profile

▼

Per-period learned baselines (EMA). Set count≥10 for confidence. Leave at 0 to use fallback chain.

Exercise & Sleep Modifiers

▼

Exercise Sessions (last 48h)

Duration (min)

Hours Ago

Sleep (last 24h)

Total Sleep Hours

Typical = 7.5h. Deficit raises dawn/morning baseline. Only active during dawn (5-8) and morning (8-11) periods.

Personalization Factors

▼

Carb Sensitivity

Baseline Offset

Peak Time Offset

Meals Used

Protein Blunting

Fiber Blunting

Personalization activates when meals used ≥ 3. Confidence = meals_used / 30.

Learning Feedback Simulator

▼

Simulate CGM outcome to see how factors update (mirrors update_factors_from_outcome).

Actual Rise (mg/dL)

Actual Time to Peak (min)

Excursion Output

-- mg/dL

Predicted Peak

-- mg/dL

Predicted Rise

-- min

Time to Peak

-- mg/dL

Adj. Baseline

-- %

Confidence

-- mg/dL

At 2h

-- mg/dL

Exercise Offset

-- mg/dL

Sleep Offset

GR:eat Prediction Tester

Full offline replica — circadian, exercise, sleep, meal stacking

Meal Presets

Current Meal Macros

Meal Stacking (Active Meals)

User Profile

Circadian Glucose Profile

Exercise & Sleep Modifiers

Exercise Sessions (last 48h)

Sleep (last 24h)

Personalization Factors

Learning Feedback Simulator

Excursion Output

Glucose Excursion Curve (4 hours)

Prediction Breakdown