ddmitry/clickstream-ch-kafka-superset-demo
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects Releases Wiki Activity

fix(superset): fix dashboard chart rendering in Superset 4

Browse Source 
- Why:
  - dashboard tiles failed with "Item with key 'echarts_bar' is not registered".
  - existing slice query_context stayed stale after config updates.
- What:
  - switch Top Pages and Data Quality Summary from \'echarts_bar\' to \'dist_bar\'.
  - use \'groupby\' for categorical bar charts and sync this into query_context.
  - keep dashboard export config aligned with runtime chart definitions.
- Check:
  - python3 -m py_compile superset/create_dashboard.py
  - docker compose exec -T superset python /app/superset_init/create_dashboard.py
  - DB check for slices 9/10: viz_type=form_data=query_context set to dist_bar
This commit is contained in:
ddadmin
2026-02-10 23:49:40 +03:00
parent cb3665c1be
commit 6533f8b32c
2 changed files with 206 additions and 85 deletions
Download Patch File Download Diff File Expand all files Collapse all files
superset/create_dashboard.py
+178 -57
View File
@@ -5,7 +5,7 @@
================================================================================
Назначение:
- Создание чартов (Charts) на основе датасетов DM-слоя
- Создание дашборда с layout
- Создание дашборда с layout и native filters
Запуск:
Внутри контейнера superset:
@@ -41,6 +41,7 @@ CHARTS_CONFIG = [
"label": "Total Events",
"optionName": "metric_1"
},
"granularity_sqla": "event_ts",
"y_axis_format": ",d",
"show_trend_line": False,
"time_range": "No filter"
@@ -57,6 +58,7 @@ CHARTS_CONFIG = [
"label": "Unique Users",
"optionName": "metric_2"
},
"granularity_sqla": "event_ts",
"y_axis_format": ",d",
"show_trend_line": False,
"time_range": "No filter"
@@ -73,6 +75,7 @@ CHARTS_CONFIG = [
"label": "Unique Sessions",
"optionName": "metric_3"
},
"granularity_sqla": "event_ts",
"y_axis_format": ",d",
"show_trend_line": False,
"time_range": "No filter"
@@ -89,6 +92,7 @@ CHARTS_CONFIG = [
"label": "Avg Events/Session",
"optionName": "metric_4"
},
"granularity_sqla": "event_ts",
"y_axis_format": ".2f",
"show_trend_line": False,
"time_range": "No filter"
@@ -178,15 +182,14 @@ CHARTS_CONFIG = [
},
{
"slice_name": "📄 Top Pages",
"viz_type": "echarts_bar",
"viz_type": "dist_bar",
"dataset_name": "v_top_pages_daily",
"params": {
"x_axis": "page_url_path",
"groupby": ["page_url_path"],
"metrics": [
{"expressionType": "SQL", "sqlExpression": "SUM(pageviews)", "label": "Pageviews"}
],
"row_limit": 20,
"order_by_cols": [["SUM(pageviews)", False]],
"time_range": "No filter",
"orientation": "vertical",
"show_legend": False
@@ -195,10 +198,10 @@ CHARTS_CONFIG = [
# Качество данных
{
"slice_name": "🔍 Data Quality Summary",
"viz_type": "echarts_bar",
"viz_type": "dist_bar",
"dataset_name": "dq_summary",
"params": {
"x_axis": "layer",
"groupby": ["layer"],
"metrics": [
{"expressionType": "SQL", "sqlExpression": "SUM(check_value)", "label": "Row Count"}
],
@@ -225,15 +228,82 @@ DASHBOARD_CONFIG = {
"description": "Аналитический дашборд для e-commerce кликстрима: трафик, конверсии, география и качество данных.",
"published": True,
"slug": "ecommerce-analytics",
"json_metadata": json.dumps({
"native_filter_configuration": [
}
def sync_query_context(chart, params: dict, dataset_id: int) -> None:
"""
Синхронизирует сохраненный query_context с обновленными params чарта.
Для Superset 4.x у `big_number` запрос валидируется как time-series и
ожидает `granularity` в query_context (проверено по актуальной документации).
"""
if not chart.query_context:
return
try:
query_context = json.loads(chart.query_context)
except (TypeError, json.JSONDecodeError):
logger.warning("Chart ID %s has invalid query_context, skip sync", chart.id)
return
query_context["datasource"] = {"id": dataset_id, "type": "table"}
query_context["form_data"] = {
**params,
"datasource": f"{dataset_id}__table",
"viz_type": chart.viz_type,
"slice_id": chart.id,
}
queries = query_context.get("queries")
if not isinstance(queries, list) or not queries:
chart.query_context = json.dumps(query_context)
return
if chart.viz_type == "big_number":
query = queries[0]
granularity = params.get("granularity_sqla")
if granularity:
query["granularity"] = granularity
query["is_timeseries"] = True
query["time_range"] = params.get("time_range", query.get("time_range"))
if "metric" in params:
query["metrics"] = [params["metric"]]
extras = query.get("extras") if isinstance(query.get("extras"), dict) else {}
if "time_grain_sqla" in params:
extras["time_grain_sqla"] = params.get("time_grain_sqla")
query["extras"] = extras
elif params.get("x_axis") or params.get("groupby"):
# Для категориальных графиков синхронизируем колонки измерений.
dimensions = params.get("groupby")
if not dimensions and params.get("x_axis"):
dimensions = [params["x_axis"]]
query = queries[0]
query["columns"] = dimensions
if "metrics" in params:
query["metrics"] = params["metrics"]
elif "metric" in params:
query["metrics"] = [params["metric"]]
query["row_limit"] = params.get("row_limit", query.get("row_limit"))
query["time_range"] = params.get("time_range", query.get("time_range"))
query["is_timeseries"] = False
chart.query_context = json.dumps(query_context)
def build_dashboard_metadata(filter_dataset_id: int | None) -> str:
"""Формирует json_metadata с валидными datasetId для native filters."""
native_filters = []
if filter_dataset_id is not None:
native_filters = [
{
"id": "date_filter",
"name": "📅 Date Range",
"filterType": "filter_time",
"targets": [{"datasetId": None, "column": {"name": "event_date"}}],
"targets": [{"datasetId": filter_dataset_id, "column": {"name": "event_date"}}],
"defaultValue": "Last week",
"scope": {"root": ["ROOT_ID"], "excluded": []},
"scope": {"rootPath": ["ROOT_ID"], "excluded": []},
"cascadeParentIds": [],
"isInstant": True
},
@@ -241,8 +311,8 @@ DASHBOARD_CONFIG = {
"id": "country_filter",
"name": "🌍 Country",
"filterType": "filter_select",
"targets": [{"datasetId": None, "column": {"name": "geo_country"}}],
"scope": {"root": ["ROOT_ID"], "excluded": []},
"targets": [{"datasetId": filter_dataset_id, "column": {"name": "geo_country"}}],
"scope": {"rootPath": ["ROOT_ID"], "excluded": []},
"isInstant": True,
"allowsMultipleValues": True,
"isRequired": False
@@ -251,8 +321,8 @@ DASHBOARD_CONFIG = {
"id": "device_filter",
"name": "📱 Device Type",
"filterType": "filter_select",
"targets": [{"datasetId": None, "column": {"name": "device_type"}}],
"scope": {"root": ["ROOT_ID"], "excluded": []},
"targets": [{"datasetId": filter_dataset_id, "column": {"name": "device_type"}}],
"scope": {"rootPath": ["ROOT_ID"], "excluded": []},
"isInstant": True,
"allowsMultipleValues": True,
"isRequired": False
@@ -261,20 +331,23 @@ DASHBOARD_CONFIG = {
"id": "browser_filter",
"name": "🌐 Browser",
"filterType": "filter_select",
"targets": [{"datasetId": None, "column": {"name": "browser_name"}}],
"scope": {"root": ["ROOT_ID"], "excluded": []},
"targets": [{"datasetId": filter_dataset_id, "column": {"name": "browser_name"}}],
"scope": {"rootPath": ["ROOT_ID"], "excluded": []},
"isInstant": True,
"allowsMultipleValues": True,
"isRequired": False
}
],
]
metadata = {
"native_filter_configuration": native_filters,
"color_scheme": "supersetColors",
"label_colors": {}
})
}
}
return json.dumps(metadata)
def main():
def main() -> bool:
"""Главная функция"""
logger.info("=" * 60)
logger.info("Creating E-commerce Analytics Dashboard")
@@ -292,6 +365,7 @@ def main():
from superset.connectors.sqla.models import SqlaTable
created_charts = []
datasets_by_name = {}
# Создаём чарты
for chart_config in CHARTS_CONFIG:
@@ -303,23 +377,37 @@ def main():
if not dataset:
logger.warning(f"Dataset '{chart_config['dataset_name']}' not found, skipping chart")
continue
datasets_by_name[chart_config["dataset_name"]] = dataset.id
try:
# Подготавливаем параметры
params = chart_config["params"].copy()
params["datasource"] = f"{dataset.id}__table"
params["viz_type"] = chart_config["viz_type"]
serialized_params = json.dumps(params)
# Проверяем, существует ли уже чарт
existing = db.session.query(Slice).filter_by(
slice_name=chart_config["slice_name"]
).first()
if existing:
logger.info(f"Chart '{chart_config['slice_name']}' already exists (ID: {existing.id})")
# Синхронизируем параметры существующего чарта с конфигом.
existing.viz_type = chart_config["viz_type"]
existing.datasource_id = dataset.id
existing.datasource_type = "table"
existing.datasource_name = dataset.table_name
existing.params = serialized_params
sync_query_context(existing, params, dataset.id)
existing.description = f"Chart created automatically for {chart_config['dataset_name']}"
db.session.flush()
logger.info(
f"Chart '{chart_config['slice_name']}' already exists (ID: {existing.id}), "
"params synced"
)
created_charts.append({"id": existing.id, "title": existing.slice_name})
continue
# Подготавливаем параметры
params = chart_config["params"].copy()
params["datasource"] = f"{dataset.id}__table"
params["viz_type"] = chart_config["viz_type"]
# Создаём чарт
chart = Slice(
slice_name=chart_config["slice_name"],
@@ -327,7 +415,7 @@ def main():
datasource_id=dataset.id,
datasource_type="table",
datasource_name=dataset.table_name,
params=json.dumps(params),
params=serialized_params,
description=f"Chart created automatically for {chart_config['dataset_name']}"
)
@@ -344,6 +432,51 @@ def main():
db.session.rollback()
logger.info(f"Created/Found {len(created_charts)} charts")
metadata_json = build_dashboard_metadata(datasets_by_name.get("v_events_enriched"))
# Создаём позиции чартов для layout.
# Обязательные блоки ROOT_ID/GRID_ID нужны для корректной работы /tabs.
positions = {
"DASHBOARD_VERSION_KEY": "v2",
"ROOT_ID": {
"id": "ROOT_ID",
"type": "ROOT",
"children": ["GRID_ID"],
},
"GRID_ID": {
"id": "GRID_ID",
"type": "GRID",
"children": [],
"parents": ["ROOT_ID"],
"meta": {"background": "BACKGROUND_TRANSPARENT"},
},
}
# Добавляем чарты в layout (grid: 12 columns)
y_position = 0
chart_index = 0
for chart in created_charts:
if chart:
chart_component_id = f"CHART-{chart['id']}"
positions[chart_component_id] = {
"id": chart_component_id,
"type": "CHART",
"children": [],
"parents": ["ROOT_ID", "GRID_ID"],
"meta": {
"chartId": chart['id'],
"sliceName": chart['title'],
"height": 50,
"width": 4 if chart_index < 4 else 6,
"x": (chart_index % 3) * 4 if chart_index < 4 else (chart_index % 2) * 6,
"y": y_position,
},
}
positions["GRID_ID"]["children"].append(chart_component_id)
chart_index += 1
if chart_index % 4 == 0:
y_position += 50
# Создаём дашборд
if created_charts:
@@ -354,38 +487,22 @@ def main():
).first()
if existing:
existing.description = DASHBOARD_CONFIG["description"]
existing.published = DASHBOARD_CONFIG["published"]
existing.json_metadata = metadata_json
existing.position_json = json.dumps(positions)
existing.slices = []
for chart_info in created_charts:
chart = db.session.query(Slice).filter_by(id=chart_info["id"]).first()
if chart:
existing.slices.append(chart)
db.session.commit()
logger.info(f"Dashboard '{DASHBOARD_CONFIG['dashboard_title']}' already exists (ID: {existing.id})")
logger.info("=" * 60)
logger.info("Dashboard already exists!")
logger.info("Dashboard already exists and metadata/layout were updated.")
logger.info(f"Dashboard URL: /superset/dashboard/{existing.id}/")
logger.info("=" * 60)
return
# Создаём позиции чартов для layout
positions = {"DASHBOARD_VERSION_KEY": "v2"}
# Добавляем чарты в layout (grid: 12 columns)
y_position = 0
chart_index = 0
for chart in created_charts:
if chart:
positions[f"CHART-{chart['id']}"] = {
"id": f"CHART-{chart['id']}",
"type": "CHART",
"parents": ["ROOT_ID"],
"meta": {
"chartId": chart['id'],
"sliceName": chart['title'],
"height": 50,
"width": 4 if chart_index < 4 else 6,
"x": (chart_index % 3) * 4 if chart_index < 4 else (chart_index % 2) * 6,
"y": y_position
}
}
chart_index += 1
if chart_index % 4 == 0:
y_position += 50
return True
# Создаём дашборд
dashboard = Dashboard(
@@ -393,7 +510,7 @@ def main():
slug=DASHBOARD_CONFIG["slug"],
description=DASHBOARD_CONFIG["description"],
published=DASHBOARD_CONFIG["published"],
json_metadata=DASHBOARD_CONFIG["json_metadata"],
json_metadata=metadata_json,
position_json=json.dumps(positions)
)
@@ -414,15 +531,19 @@ def main():
logger.info("Dashboard created successfully!")
logger.info(f"Dashboard URL: /superset/dashboard/{dashboard.id}/")
logger.info("=" * 60)
return True
except Exception as e:
logger.error(f"Failed to create dashboard: {e}")
import traceback
traceback.print_exc()
db.session.rollback()
return False
else:
logger.error("No charts created, cannot create dashboard")
return False
return False
if __name__ == "__main__":
main()
sys.exit(0 if main() else 1)
superset/dashboards/ecommerce_analytics.zip.json
+4 -4
View File
@@ -96,20 +96,20 @@
{
"__Slice__": {
"slice_name": "📄 Top Pages",
"viz_type": "echarts_bar",
"viz_type": "dist_bar",
"datasource_type": "table",
"datasource_name": "dm.v_top_pages_daily",
"params": "{\"datasource\": \"3__table\", \"viz_type\": \"echarts_bar\", \"x_axis\": \"page_url_path\", \"metrics\": [{\"expressionType\": \"SQL\", \"sqlExpression\": \"SUM(pageviews)\", \"label\": \"Pageviews\"}], \"row_limit\": 20, \"order_by_cols\": [[\"SUM(pageviews)\", false]], \"time_range\": \"No filter\", \"orientation\": \"vertical\", \"show_legend\": false}",
"params": "{\"datasource\": \"3__table\", \"viz_type\": \"dist_bar\", \"groupby\": [\"page_url_path\"], \"metrics\": [{\"expressionType\": \"SQL\", \"sqlExpression\": \"SUM(pageviews)\", \"label\": \"Pageviews\"}], \"row_limit\": 20, \"order_by_cols\": [[\"SUM(pageviews)\", false]], \"time_range\": \"No filter\", \"orientation\": \"vertical\", \"show_legend\": false}",
"description": "Топ страниц по просмотрам"
}
},
{
"__Slice__": {
"slice_name": "🔍 Data Quality Summary",
"viz_type": "echarts_bar",
"viz_type": "dist_bar",
"datasource_type": "table",
"datasource_name": "dm.dq_summary",
"params": "{\"datasource\": \"4__table\", \"viz_type\": \"echarts_bar\", \"x_axis\": \"layer\", \"metrics\": [{\"expressionType\": \"SQL\", \"sqlExpression\": \"SUM(check_value)\", \"label\": \"Row Count\"}], \"adhoc_filters\": [{\"clause\": \"WHERE\", \"expressionType\": \"SQL\", \"sqlExpression\": \"check_name = 'total_rows'\", \"subject\": null, \"operator\": null, \"comparator\": null}], \"row_limit\": 100, \"time_range\": \"No filter\", \"show_legend\": false}",
"params": "{\"datasource\": \"4__table\", \"viz_type\": \"dist_bar\", \"groupby\": [\"layer\"], \"metrics\": [{\"expressionType\": \"SQL\", \"sqlExpression\": \"SUM(check_value)\", \"label\": \"Row Count\"}], \"adhoc_filters\": [{\"clause\": \"WHERE\", \"expressionType\": \"SQL\", \"sqlExpression\": \"check_name = 'total_rows'\", \"subject\": null, \"operator\": null, \"comparator\": null}], \"row_limit\": 100, \"time_range\": \"No filter\", \"show_legend\": false}",
"description": "Сводка по качеству данных"
}
}