using QuestionsApi.Models;

namespace QuestionsApi.Services;

public static class ExamSimulationScoring
{
    public const string ModelVersion = "provisional-2";
    private const double PassingAccuracy = 0.60;
    private const double ScoreIntercept = 50.0;
    private const double ScoreScale = 250.0;

    public static ExamSimulationScore Calculate(
        IReadOnlyList<ExamSimulationItem> items,
        IReadOnlyDictionary<string, ExamSimulationResponse> responses,
        IReadOnlyDictionary<string, Question> questions)
    {
        var correct = 0;
        var seen = 0;
        var weightedCorrect = 0.0;
        var weightSum = 0.0;
        var squaredWeightSum = 0.0;
        var categories = new Dictionary<string, ExamSimulationCategoryScore>();

        foreach (var item in items)
        {
            questions.TryGetValue(item.QuestionExternalId, out var question);
            responses.TryGetValue(item.QuestionExternalId, out var response);
            var isCorrect = question is not null && response?.ChoiceId == question.CorrectChoiceId;
            if (isCorrect) correct++;
            if (item.WasSeen) seen++;
            var difficulty = ResolveDifficulty(item, question);
            var weight = DifficultyWeight(difficulty);
            weightSum += weight;
            squaredWeightSum += weight * weight;
            if (isCorrect) weightedCorrect += weight;
            if (!categories.TryGetValue(item.BlueprintCategory, out var category))
            {
                category = new ExamSimulationCategoryScore();
                categories[item.BlueprintCategory] = category;
            }
            category.Total++;
            if (isCorrect) category.Correct++;
        }

        var rawAccuracy = items.Count == 0 ? 0.0 : (double)correct / items.Count;
        var adjustedAccuracy = weightSum == 0.0 ? 0.0 : weightedCorrect / weightSum;
        var predicted = ToScore(adjustedAccuracy);
        var effectiveSampleSize = squaredWeightSum == 0.0 ? 0.0 : weightSum * weightSum / squaredWeightSum;
        var accuracySe = effectiveSampleSize == 0.0
            ? 0.0
            : Math.Sqrt(adjustedAccuracy * (1.0 - adjustedAccuracy) / effectiveSampleSize);
        var accuracyMargin = 1.96 * accuracySe;
        var rangeLow = ToScore(Math.Clamp(adjustedAccuracy - accuracyMargin, 0.0, 1.0));
        var rangeHigh = ToScore(Math.Clamp(adjustedAccuracy + accuracyMargin, 0.0, 1.0));
        var passProbability = accuracySe == 0.0
            ? adjustedAccuracy >= PassingAccuracy ? 1.0 : 0.0
            : NormalCdf((adjustedAccuracy - PassingAccuracy) / accuracySe);
        return new ExamSimulationScore
        {
            PredictedScore = predicted,
            RangeLow = rangeLow,
            RangeHigh = rangeHigh,
            PassProbability = Math.Round(passProbability, 4),
            Correct = correct,
            Total = items.Count,
            SeenCount = seen,
            RawAccuracy = Math.Round(rawAccuracy, 4),
            DifficultyAdjustedAccuracy = Math.Round(adjustedAccuracy, 4),
            AbilityElo = null,
            Categories = categories
        };
    }

    public static int ResolveDifficulty(ExamSimulationItem item, Question? question = null)
    {
        if (item.QuestionDifficulty is >= 1 and <= 5) return item.QuestionDifficulty;
        return question?.Difficulty is >= 1 and <= 5 ? question.Difficulty : 3;
    }

    private static double DifficultyWeight(int difficulty) => difficulty switch
    {
        1 => 0.8,
        2 => 0.9,
        3 => 1.0,
        4 => 1.1,
        5 => 1.2,
        _ => 1.0
    };

    private static int ToScore(double accuracy) =>
        Math.Clamp((int)Math.Round(ScoreIntercept + ScoreScale * accuracy), 1, 300);

    private static double NormalCdf(double value) => 0.5 * (1.0 + Erf(value / Math.Sqrt(2.0)));

    private static double Erf(double value)
    {
        var sign = value < 0 ? -1.0 : 1.0;
        value = Math.Abs(value);
        const double p = 0.3275911;
        var t = 1.0 / (1.0 + p * value);
        var y = 1.0 - (((((1.061405429 * t - 1.453152027) * t) + 1.421413741) * t - 0.284496736) * t + 0.254829592) * t * Math.Exp(-value * value);
        return sign * y;
    }
}