trading_bot_v3/app/api/ai-analytics/route.js

import { NextResponse } from 'next/server'
import { PrismaClient } from '@prisma/client'

const prisma = new PrismaClient()

export async function GET() {
  try {
    console.log('🔍 AI Analytics API called')
    
    // Calculate date range for analytics (last 30 days)
    const endDate = new Date()
    const startDate = new Date()
    startDate.setDate(startDate.getDate() - 30)
    
    // Get learning data using correct snake_case model name
    const learningData = await prisma.ai_learning_data.findMany({
      where: {
        createdAt: {
          gte: startDate
        }
      },
      orderBy: { createdAt: 'desc' },
      take: 1000
    })

    // Get trade data
    const trades = await prisma.trades.findMany({
      where: {
        createdAt: {
          gte: startDate
        }
      },
      orderBy: { createdAt: 'desc' },
      take: 100
    })

    // Get automation sessions
    const sessions = await prisma.automation_sessions.findMany({
      where: {
        createdAt: {
          gte: startDate
        }
      },
      orderBy: { createdAt: 'desc' },
      take: 50
    })

    // Calculate analytics
    const overview = {
      totalLearningRecords: learningData.length,
      totalTrades: trades.length,
      totalSessions: sessions.length,
      activeSessions: sessions.filter(s => s.status === 'ACTIVE').length
    }

    // Calculate improvements
    const recentData = learningData.slice(0, Math.floor(learningData.length / 2))
    const olderData = learningData.slice(Math.floor(learningData.length / 2))
    
    const recentAvgConfidence = recentData.length > 0 
      ? recentData.reduce((sum, d) => sum + (d.confidenceScore || 50), 0) / recentData.length 
      : 50
    const olderAvgConfidence = olderData.length > 0 
      ? olderData.reduce((sum, d) => sum + (d.confidenceScore || 50), 0) / olderData.length 
      : 50

    const improvements = {
      confidenceImprovement: recentAvgConfidence - olderAvgConfidence,
      accuracyImprovement: null, // Would need actual outcome tracking
      trend: recentAvgConfidence > olderAvgConfidence ? 'improving' : 'declining'
    }

    // Calculate P&L from trades
    const completedTrades = trades.filter(t => t.status === 'COMPLETED')
    const totalPnL = completedTrades.reduce((sum, t) => sum + (t.profit || 0), 0)
    const winningTrades = completedTrades.filter(t => (t.profit || 0) > 0)
    const winRate = completedTrades.length > 0 ? winningTrades.length / completedTrades.length : 0

    const pnl = {
      totalTrades: completedTrades.length,
      totalPnL: totalPnL,
      totalPnLPercent: 0, // Would need to calculate based on initial capital
      winRate: winRate,
      avgTradeSize: completedTrades.length > 0 
        ? completedTrades.reduce((sum, t) => sum + (t.amount || 0), 0) / completedTrades.length 
        : 0
    }

    // Get current position (if any)
    const latestTrade = trades.find(t => t.status === 'ACTIVE' || t.status === 'PENDING')
    const currentPosition = latestTrade ? {
      symbol: latestTrade.symbol,
      side: latestTrade.side,
      amount: latestTrade.amount,
      entryPrice: latestTrade.price,
      unrealizedPnL: 0 // Would need current market price to calculate
    } : null

    // Real-time metrics
    const firstLearningRecord = learningData[learningData.length - 1]
    const daysSinceStart = firstLearningRecord 
      ? Math.ceil((Date.now() - new Date(firstLearningRecord.createdAt).getTime()) / (1000 * 60 * 60 * 24))
      : 0

    const realTimeMetrics = {
      daysSinceAIStarted: daysSinceStart,
      learningRecordsPerDay: daysSinceStart > 0 ? learningData.length / daysSinceStart : 0,
      tradesPerDay: daysSinceStart > 0 ? trades.length / daysSinceStart : 0,
      lastUpdate: new Date().toISOString(),
      isLearningActive: sessions.some(s => s.status === 'ACTIVE')
    }

    // Learning proof
    const learningProof = {
      hasImprovement: improvements.confidenceImprovement > 0,
      improvementDirection: improvements.trend,
      confidenceChange: improvements.confidenceImprovement,
      sampleSize: learningData.length,
      isStatisticallySignificant: learningData.length > 50 && Math.abs(improvements.confidenceImprovement) > 5
    }

    const analytics = {
      generated: new Date().toISOString(),
      overview,
      improvements,
      pnl,
      currentPosition,
      realTimeMetrics,
      learningProof
    }

    console.log('✅ AI Analytics generated successfully')
    return NextResponse.json({
      success: true,
      data: analytics
    })

  } catch (error) {
    console.error('Error generating AI analytics:', error)
    return NextResponse.json({
      success: false,
      error: 'Failed to generate AI analytics',
      details: error.message
    }, { status: 500 })
  }
}