POST /v1/reports/path_to_conversion?project_id={project_id}

{
  "ResponseType": "JSON",
  "Fields": [
    "InteractionTime",
    "InteractionType",
    "MediaCampaignName",
    "TargetEventName"
  ],
  "InteractionFilter": {
    "DateFrom": "2024-01-01",
    "DateTo": "2024-01-31"
  },
  "TargetFilter": {
    "DateFrom": "2024-01-01",
    "DateTo": "2024-01-31",
    "EventType": ["Purchase"]
  }
}

curl -X POST "https://api.targetads.io/v1/reports/path_to_conversion?project_id=12486" \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "ResponseType": "JSON",
    "Fields": [
      "PathDeviceID",
      "PathPosition",
      "PathLength",
      "InteractionTime",
      "InteractionType",
      "InteractionMediaCampaignName",
      "TargetEventName",
      "TargetEcomAmount",
      "Weight_MLI",
      "Weight_FL"
    ],
    "TargetFilter": {
      "DateFrom": "2024-01-01",
      "DateTo": "2024-01-31",
      "EventType": ["Purchase"]
    },
    "Limit": 10000
  }'

curl -X POST "https://api.targetads.io/v1/reports/path_to_conversion?project_id=12486" \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "ResponseType": "JSON",
    "Fields": [
      "PathDeviceID",
      "PathPosition",
      "InteractionTime",
      "InteractionType",
      "InteractionMediaCampaignName",
      "InteractionUtmSource",
      "TargetEventName",
      "Weight_MLI",
      "Weight_MFI",
      "Weight_ML"
    ],
    "InteractionFilter": {
      "MediaCampaignId": [12345, 12346]
    },
    "TargetFilter": {
      "DateFrom": "2024-01-01",
      "DateTo": "2024-01-31",
      "EventType": ["Purchase"]
    }
  }'

curl -X POST "https://api.targetads.io/v1/reports/path_to_conversion?project_id=12486" \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "ResponseType": "JSON",
    "Fields": [
      "PathDeviceID",
      "PathPosition",
      "PathLength",
      "InteractionTime",
      "InteractionType",
      "InteractionMediaCampaignName",
      "TargetEcomId",
      "TargetEcomAmount",
      "TargetEcomItemsName",
      "TargetEcomItemsCategory1",
      "Weight_FTDM_30"
    ],
    "TargetFilter": {
      "DateFrom": "2024-01-01",
      "DateTo": "2024-01-31",
      "EventType": ["Purchase"],
      "EcomItemCategory1": ["Electronics", "Books"]
    }
  }'

curl -X POST "https://api.targetads.io/v1/reports/path_to_conversion?project_id=12486" \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "ResponseType": "JSON",
    "Fields": [
      "PathPosition",
      "InteractionTime",
      "InteractionType",
      "InteractionUtmSource",
      "InteractionUtmMedium",
      "InteractionUtmCampaign",
      "TargetEventName",
      "Weight_FL"
    ],
    "InteractionFilter": {
      "UtmSource": ["google", "facebook"]
    },
    "TargetFilter": {
      "DateFrom": "2024-01-01",
      "DateTo": "2024-01-31"
    }
  }'

{
  "data": [
    {
      "PathDeviceID": "abc123",
      "PathPosition": "1",
      "PathLength": "4",
      "InteractionTime": "2024-01-10 10:00:00",
      "InteractionType": "Impression",
      "InteractionMediaCampaignName": "Campaign A",
      "TargetEventName": "Purchase",
      "TargetEcomAmount": "1500.00",
      "Weight_MLI": "0",
      "Weight_FL": "0.25"
    },
    {
      "PathDeviceID": "abc123",
      "PathPosition": "2",
      "PathLength": "4",
      "InteractionTime": "2024-01-12 14:30:00",
      "InteractionType": "Click",
      "InteractionMediaCampaignName": "Campaign B",
      "TargetEventName": "Purchase",
      "TargetEcomAmount": "1500.00",
      "Weight_MLI": "0",
      "Weight_FL": "0.25"
    },
    {
      "PathDeviceID": "abc123",
      "PathPosition": "3",
      "PathLength": "4",
      "InteractionTime": "2024-01-14 16:00:00",
      "InteractionType": "PageView",
      "InteractionMediaCampaignName": "",
      "TargetEventName": "Purchase",
      "TargetEcomAmount": "1500.00",
      "Weight_MLI": "0",
      "Weight_FL": "0.25"
    },
    {
      "PathDeviceID": "abc123",
      "PathPosition": "4",
      "PathLength": "4",
      "InteractionTime": "2024-01-15 18:20:00",
      "InteractionType": "Click",
      "InteractionMediaCampaignName": "Campaign C",
      "TargetEventName": "Purchase",
      "TargetEcomAmount": "1500.00",
      "Weight_MLI": "1",
      "Weight_FL": "0.25"
    }
  ],
  "count": 4
}

# Группировка по путям
paths = {}
for row in data:
    device_id = row['PathDeviceID']
    if device_id not in paths:
        paths[device_id] = []
    paths[device_id].append(row)

# Анализ каждого пути
for device_id, touchpoints in paths.items():
    path_length = touchpoints[0]['PathLength']
    conversion_value = touchpoints[0]['TargetEcomAmount']
    print(f"Device {device_id}: {path_length} touchpoints, value {conversion_value}")

def calculate_campaign_contribution(data, weight_field='Weight_MLI'):
    """Расчет вклада каждой кампании в конверсии"""
    campaign_stats = {}

    for row in data:
        campaign = row['InteractionMediaCampaignName']
        if not campaign:
            continue

        if campaign not in campaign_stats:
            campaign_stats[campaign] = {
                'conversions': 0,
                'revenue': 0
            }

        weight = float(row[weight_field])
        revenue = float(row['TargetEcomAmount'])

        campaign_stats[campaign]['conversions'] += weight
        campaign_stats[campaign]['revenue'] += weight * revenue

    return campaign_stats

def analyze_path_lengths(data):
    """Анализ распределения длины путей"""
    path_lengths = {}

    for row in data:
        if row['PathPosition'] == '1':  # Только первое касание
            length = int(row['PathLength'])
            path_lengths[length] = path_lengths.get(length, 0) + 1

    return path_lengths

# Результат: {1: 150, 2: 340, 3: 280, 4: 190, 5: 40}
# 150 конверсий с одним касанием, 340 с двумя и т.д.

def analyze_campaign_positions(data):
    """Где в пути чаще встречаются кампании"""
    campaign_positions = {}

    for row in data:
        campaign = row['InteractionMediaCampaignName']
        if not campaign:
            continue

        position = int(row['PathPosition'])

        if campaign not in campaign_positions:
            campaign_positions[campaign] = {}

        campaign_positions[campaign][position] = \
            campaign_positions[campaign].get(position, 0) + 1

    return campaign_positions

def get_all_paths(project_id, date_from, date_to):
    offset = 0
    limit = 100000
    all_paths = []

    while True:
        response = get_path_to_conversion(
            project_id=project_id,
            date_from=date_from,
            date_to=date_to,
            offset=offset,
            limit=limit
        )

        all_paths.extend(response['data'])

        if len(response['data']) < limit:
            break

        offset += limit

    return all_paths

# Отдельно для разных типов конверсий
for event_type in ['Lead', 'Purchase', 'Registration']:
    paths = get_path_to_conversion(
        target_filter={'EventType': [event_type]}
    )
    analyze_paths(paths, event_type)

# Минимальный набор для анализа атрибуции
fields = [
    'PathDeviceID',
    'PathPosition',
    'InteractionTime',
    'InteractionMediaCampaignName',
    'TargetEcomAmount',
    'Weight_MLI'
]

# Запросить данные со всеми весами
fields = [
    'PathDeviceID',
    'InteractionMediaCampaignName',
    'TargetEcomAmount',
    'Weight_MFI',
    'Weight_MLI',
    'Weight_ML',
    'Weight_FL'
]

# Рассчитать вклад по каждой модели
for model in ['Weight_MFI', 'Weight_MLI', 'Weight_ML', 'Weight_FL']:
    contribution = calculate_campaign_contribution(data, model)
    print(f"{model}: {contribution}")

def analyze_journey_patterns(data):
    """Найти популярные последовательности касаний"""
    patterns = {}

    paths = group_by_device(data)

    for device_id, touchpoints in paths.items():
        pattern = ' -> '.join([
            f"{t['InteractionType']}({t['InteractionMediaCampaignName']})"
            for t in touchpoints
        ])
        patterns[pattern] = patterns.get(pattern, 0) + 1

    return sorted(patterns.items(), key=lambda x: x[1], reverse=True)

from datetime import datetime

def calculate_conversion_time(data):
    """Время от первого касания до конверсии"""
    paths = group_by_device(data)

    conversion_times = []

    for device_id, touchpoints in paths.items():
        first = datetime.strptime(touchpoints[0]['InteractionTime'], '%Y-%m-%d %H:%M:%S')
        last = datetime.strptime(touchpoints[-1]['InteractionTime'], '%Y-%m-%d %H:%M:%S')

        hours = (last - first).total_seconds() / 3600
        conversion_times.append(hours)

    return {
        'avg_hours': sum(conversion_times) / len(conversion_times),
        'median_hours': sorted(conversion_times)[len(conversion_times) // 2]
    }