1.1 Error messages #

What should I do if the dataset involved in the subquery does not have read permission? #

Cause
A subquery filter is used in the visualization query, and the chart for the subquery does not have permission.
Solution
Request view permission for the dataset corresponding to the subquery chart.

What should I do if the number of data points exceeds the limit? #

Cause
The number of data points rendered in the chart exceeds the rendering limit.
Solution
(1) Use a table to display the data.
(2) If the x-axis is set to a detailed time granularity, click the time field in the dimension configuration bar to change the aggregation method. It is recommended to change the aggregation to hour, day, or month based on actual display requirements.

What should I do if an unsupported filter value error occurs? #

Cause
There is an abnormal filter, which may be due to a mismatch between the filter field and the filter value type.
Solution
Check the filters in the filter bar. If an abnormal filter is found (for example, a text-type field uses a filter such as ">0"), remove it and reconfigure the filter condition by dragging and dropping.

What should I do if the fields used by the chart have been deleted? #

Cause
Some fields used by the chart have been deleted.
Solution
Contact the dataset owner to confirm whether there are replacement fields and reconfigure.

When configuring a map, an error occurs stating that the dimension field used by the chart must be set to a geographic role. #

Cause
The dimension field used is not a geographic role field and needs to be matched to a geographic role.
Solution
Match the dimension field to a geographic role (editing permission for the queried dataset is required). For detailed instructions, see 3.1 Edit geographic roles.

When configuring year-over-year or period-over-period comparison, an error occurs stating that the time range is set to unlimited, which does not support year-over-year or period-over-period comparison. #

Cause
The year-over-year or period-over-period comparison is based on the date in the filter bar, and either the start or end time is set to "unlimited". In this case, year-over-year or period-over-period comparison cannot be performed.
Solution
In the filter bar, modify the time range for the date used as the basis for period-over-period and year-over-year comparisons, and set the start and end times (these can be dynamic, but cannot be unlimited).

When configuring period-over-period or year-over-year comparison, an error occurs: period-over-period comparison is not supported for non-aggregated datetime types. #

Cause
The date used as the basis for period-over-period and year-over-year comparison is the date in the dimension bar, and it is a detailed time that has not been aggregated. In this case, period-over-period and year-over-year calculations cannot be performed.
Solution
Click the relevant dimension in the dimension bar and select a date aggregation method. It is recommended to aggregate by hour, day, or month depending on your needs.

Visualization query error: Timeout exceeded #

Possible cause
The query timed out, possibly due to cluster status or excessive data volume.
Solution
(1) Try running the query again first.
(2) Shorten the query date range to reduce the data volume.
Add filter conditions to reduce the data volume.

Visualization query error: Too many simultaneous queries #

Possible cause
Engine concurrency is too high.
Solution
(1) Try running the query again.
(2) Contact the DataWind project manager to resolve the issue.

Visualization query error: Memory limit (for query) exceeded #

Possible cause
The query data volume exceeds the limit, or the cardinality of group by / count distinct is too large, resulting in query failure.
Solution
(1) Shorten the query date range to reduce the data volume.
(2) Add filter conditions to reduce the data volume.

Visualization query error: Memory limit (for user) exceeded #

Similar issue: Visualization query error: Memory limit (total) exceeded
Possible cause
(1) The system is under heavy load and the memory limit has been exceeded.
(2) The current query requires too much memory.
Solution
(1) Try running the query again first.
(2) Shorten the query date range to reduce the data volume.
Contact the DataWind project manager to resolve the issue.

Visualization query error: Unknown identifier: field name #

Possible cause
(1) The field mentioned in the error does not exist.
(2) The table is missing some columns.
Solution
(1) Check whether the dimension or metric fields in the visualization query refer to model fields that do not exist.
(2) Contact the DataWind project manager to resolve the issue.

Visualization query error: Table...doesn't exist #

Possible cause
The table does not exist on the cluster. Possible situations:
(1) The table has been deleted.
(2) The table exists, but some nodes in the cluster are missing, resulting in an error.
Solution
(1) Check whether the underlying data table exists. If it does not exist, switch to another dataset for the query.
(2) If the table exists, contact the DataWind project manager to resolve the issue.

Visualization query error: exceeded max value of hash function #

Possible cause
There are too many dimension values in the dimension field. For example, uid
Solution
(1) Reduce the number of dimensions in the query.
(2) Shorten the query date range to reduce the amount of data queried.
(3) Add filter conditions to reduce the amount of data queried.

Visualization query error: Limit for bytes #

Possible cause
The amount of data read by the query is too large.
Solution
(1) Shorten the query date range to reduce the amount of data queried.
(2) Add filter conditions to reduce the amount of data queried.

Visualization query error: frequency reaches the upper limit #

Possible cause
There are too many complex queries. To ensure stability, the concurrency limit is triggered.
Solution
(1) Shorten the query date range to reduce the amount of data queried.
(2) Add filter conditions to reduce the amount of data queried.
(3) Contact the product project manager to resolve the issue.

Visualization query error: Could not find enough connections to up-to-date replicas #

Possible cause
There is an exception in primary and secondary cluster synchronization; the cluster is in an abnormal state.
Solution
Contact the product project manager to resolve the issue.

Visualization query error: query size is too big #

Possible cause
The SQL statement assembled by the visualization query is too long, possibly due to overly long field expressions in the query, such as excessively long case when expressions or similar expressions.
Solution
Check and adjust the field expressions.

1.2 Common issues with filters #

The date in the visualization query result does not match the date selected in the filter. #

Similar issues:

The date option does not take effect.
The time selected in the filter is inconsistent with the time of the displayed or downloaded data.

Possible cause:
The dimension date field displayed is different from the date field used as a filter, that is, the date field used for filtering and the date field used in the chart are not the same field.
Solution:
Use the business date field that needs to be filtered as the filter to select the data that actually meets the criteria.

Visualization query: filter multiple values for a single field #

Similar issues:

How to use the IN and arrayhas functions
How to filter multiple selections in visual query
How to batch upload data for filtering

When filtering, you can use IN (one value per line) or upload a file (one value per line)
For fields in array format, you can use arrayhas (separate different values with commas)

1.3 Common issues in chart analysis #

How to retrieve the top N sorted results in visual query #

Use visual analysis Top N

How to modify the data format displayed in visual query #

Click the configuration capsule for the corresponding dimension/metric, click the triangle arrow on the capsule, and select "Set Data Format"

Configure axis starting values flexibly in visual query #

In the configuration panel below the chart selection bar, click the corresponding axis to configure the axis settings.

How to add temporary custom dimensions and metrics #

Similar issues:

Fields added in visualizations are not visible to other users
Fields newly added in visual query are not visible to other users
How to view data for division, addition, operations, or aggregation of multiple metrics
How to configure division, addition, operations, or aggregation of multiple metrics

In the visual query interface, click the ⋮ next to the dataset name, select Add Field, choose to add a metric or dimension, define the display name, expression, and description to add the corresponding field. Supports aggregation, arithmetic, and other common SQL functions.
Fields added in the visualization interface are temporary and only visible to the user. To add permanent fields, dataset editing permissions are required. You can select "Save to Dataset"
or add them in dataset editing, refer to Field Configuration.

How to display different chart types for different metrics in a single chart #

Similar issues:

How to display different line styles for different metrics in a single chart

You can use a dual-axis chart to display different chart types in the same chart. How to use Dual-Axis chart
In the chart area, select "Dual-axis Chart" and configure the display method for the primary and secondary axes separately. You can also configure charts of the same type but with different line styles for the primary and secondary axes.

The data format in the line chart is inconsistent with the format set in the metric capsule. #

Similar issues:

The data format in the chart is inconsistent with the format set in the metric capsule.
The data format in the bar chart is inconsistent with the format set in the metric capsule.
The data format in the column chart is inconsistent with the format set in the metric capsule.

Please go to the chart configuration area, select the corresponding axis, and click Data Format to configure the settings.

How to adjust the data format in the axis or chart #

Similar issues:

How to adjust the data scale in the axis or chart

You can click the dimension corresponding to the value or the metric capsule, and set the data format to adjust the value format in the field. You can also select the corresponding axis (X axis or Y axis) at the bottom of the chart selection area, and click Data Format to adjust the axis coordinate format.
For details on field format and chart format configuration, see Field Configuration.

1.4 Slow query #

Visualization queries are slow, spinning for a long time without returning results #

Dashboard loading is slow
Refer to How to Optimize Query Performance? to adjust the query

1.5 Common issues with chart downloads #

No response or error when downloading visualization query data #

Similar issues:

Download failed
Failed to download data, with a timeout message
Export failed

Possible causes:
(1) Not connected to the internal network environment;
(2) Large data volume causing slow download
(3) Actual data volume is small, but the number of rows set for download is too large, causing slow download
(4) Visualization data download requires re-querying the data; if the query volume is large or there is high query concurrency, this can also cause slow downloads
Solutions:
(1) Check whether the internal network is connected;
(2) Wait patiently for 1 minute; if the data volume is large, it may take 1–2 minutes to initiate the download;
(3) Limit the number of rows to download based on the actual query volume; when the number of rows is too large, the query and download time will be extended;
(4) If the current chart's query speed is also slow, downloading data may also be slow. Please wait patiently for the download to complete.

The number of rows downloaded from the visualization query is less than the specified number of rows. #

(1) Check whether the actual amount of data is less than the number of rows to be downloaded.
(2) Check whether the top N feature is used. Disable top N and then download again.
(3) Check whether the number of rows entered during download is less than the specified number of rows. If the data exceeds 100,000 rows, it is recommended to download a CSV file.

How to set download permissions? How to control download permissions for a single chart? #

(1) Global permission settings
Users can control the download permissions for all charts within a project in the project center. This operation requires administrator permissions, and the configuration takes effect globally.
(2) Single chart permission settings
On the visualization query page, under download settings, users can set whether a chart is allowed to be downloaded in the dashboard. This operation only applies to a single chart. If the download feature for a chart is disabled, no one is allowed to download data from that chart.

1.6 Other common issues #

Inconsistent results between visualization queries and SQL queries #

Possible reasons:
(1) The query logic of the visualization query and the SQL query is inconsistent.
(2) The maximum number of rows that can be downloaded from a visualization query table is 1,000,000. If the data is not fully downloaded, the comparison may be inconsistent, but the actual data is consistent.
(3) Abnormal data in the visualization query.
Solutions:
(1) Click the query history icon in the upper left corner of the visualization query page to view the SQL details for the visualization, and compare them with the SQL query statement to confirm that the base table, field expressions, and filter conditions are all consistent.
(2) Limit the query interval and download data in segments for comparison to ensure the comparison is correct.
(3) If the query logic is consistent but the comparison results are indeed inconsistent, contact us for troubleshooting.

How to view the query SQL logic of a visualization query #

Click the query history icon in the upper left corner of the visualization query page to view the SQL statements of historical queries.

How to convert a timestamp or text-type field to a time format in visualization queries #

Click the ┇ icon in the upper right corner of the left module on the visualization page, select Batch Edit Fields, and set the field type of the corresponding field to Date or DateTime.

The most recent date with data in the visualization query is earlier than the actual most recent date with data in the dataset #

Possible causes:
(1) The page was not refreshed
(2) The successful synchronization request was lost
Solutions:
(1) Refresh the page
(2) Try to resynchronize the dataset

No data is displayed in the visualization query after the underlying table is updated #

After adding new fields to the underlying table, update the selected fields in the dataset model configuration. After updating, resynchronize the dataset so that the new fields will have data. Once the new fields for the corresponding date have data, perform the corresponding visualization query again.

No data is displayed in the query interface after removing a dimension from the visualization query #

Possible cause: Sorting or other analysis features were configured for the dimension before deletion. After deletion, analysis errors prevent display.
Solution: Click the Analysis tab next to the chart type selection area, locate the error analysis module with a red exclamation mark in the upper right corner, delete the abnormal analysis module, and apply the changes.

Where is the dashboard saved after a successful save? #

Similar questions:
How to view and manage visualization charts after saving
You can go to Project Center - Chart Management to view and centrally manage your saved charts. If they have been saved to a dashboard, you can also view them on the corresponding dashboard.

Issues with using visualization charts #

Similar questions:

How to calculate the total sum for a metric
How to display the proportion of data
How to display year-over-year and month-over-month data
How to use the comparison analysis feature
How to use the funnel chart
How to use the histogram

Please refer to the help documentation for the corresponding Table and Sorting features.

What are the requirements for using attribution analysis #

(1) The dimension must be in date format;
(2) The chart types are bar chart, column chart, line chart, or area chart. See Intelligent Attribution.

Only the most recent day's data is visible when performing a visualization query. #

Please use the p_date partition field in the 'Filter' section or select another date field to filter the query date. By default, the most recent day's data for p_date is selected when creating a visualization.

The date field format cannot be dynamically filtered. #

Similar issues:

After importing the date field into the system, it is not in date format, so it needs to be modified or filtered.

The field needs to be processed and converted to date format (dataset editing permissions are required; if you do not have permission, it is recommended to contact the dataset owner).
Method 1: In the visualization, click the upper right corner of the left sidebar, select batch edit fields, find the corresponding field, change the type to "Date" or "Datetime", and confirm.
Method 2: Hover over the corresponding field and select Edit Expression.
Use functions such as toDate to convert the field to a date or time type. Different date/time formats may require different functions. See 3. Date functions.

The derived fields for period-over-period comparison contain no data. #

Possible causes:
(1) The business data for the comparison period is missing. Example: For order date "2020-01-20", the period-over-period data for one day earlier is "2020-01-19", but the data for order date = "2020-01-19" is empty.
(2) The comparison period is missing data for the same dimension. Example: The filter condition is [Order Date] = "2020-01-20", the dimension is "City", the metric is "Sales revenue", and the period-over-period data for one day earlier is "2020-01-19". There is data for City = "Shanghai" on "2020-01-20", but no data for City = "Shanghai" on "2020-01-19", so the period-over-period calculation for "Shanghai" is empty.
Solution:
Self-check: On the period-over-period configuration page, click "View calculation rules" to check the comparison period, change the date filter condition to the corresponding date, and verify whether the business data is complete or whether the corresponding dimension item exists. Then, take appropriate action based on actual data needs, such as adjusting the comparison period or correcting the data.
Self-check example:
(1) The period-over-period calculation for 1 day earlier is empty. Reason: The dataset synchronization failed one day ago, making it impossible to retrieve the data for the corresponding period.
(2) The dimensions are "store id and store name", and the metric is "sales revenue". After performing the period-over-period comparison, the data was found to be empty. Reason: The "store id" for the comparison period did not change, but the "store name" changed, resulting in inconsistent conditions for the period-over-period comparison.