Creating a table based on an SQL query

Follow this guide to build a table based on an SQL query to the data source in Editor. For convenience, we will use the connection and dataset from the deployed demo workbook as the data source.

Getting started

To get started with DataLens, follow this deployment guide.

Deploy a demo workbook

  1. Save the dashboard titled DataLens Demo Dashboard from the Gallery:

    1. In Gallery, open the relevant creation card.
    2. Click Use → Download. The workbook will be downloaded as a JSON file.
    3. On the collections and workbooks page, click Create → Workbook.
    4. In the window that opens, click Select file and upload the JSON file with the demo workbook configuration. Optionally, edit the title and description and click Save.

  2. Navigate to the new workbook and, on the Connections tab, find 00: Demo Dashboard. Connection.

  3. Copy the connection ID by clicking → Copy ID next to it. The ID will be copied to the clipboard.

Create a chart in Editor

  1. In the workbook, click Create → Chart in Editor in the top-right corner. On the page that opens, select the Table visualization type.

  2. Link the chart with the connection by navigating to the Meta tab and adding the connection ID to links:

    {
    "links": {
        "demoConnection": "<connection_ID>"
    }
}

    Where:

    • <connection_ID>: Connection ID copied in the previous step.
    • demoConnection: Any alias name you assign to the connection and use to request chart data from the source.

    Note

    You need the Meta tab to describe service information about the list of related entities. This information is used to figure out what connections and datasets the chart is related to, as well as for the related objects dialog, when copying a workbook and when publishing to Public.

  3. Get data from the data source: to do this, open the Source tab and specify:

    module.exports = {
    'salesSourceData': {
        qlConnectionId: Editor.getId('demoConnection'),
        data: {
            sql_query: `
            SELECT
                t1."PaymentType" AS PaymentType,
                toYear(toDate(t1."OrderDatetime")) AS OrderYear,
                toStartOfInterval(toDate(toDate(t1."OrderDatetime")), interval 1 month) AS OrderMonth,
                sum(toFloat64(t1."Price")) AS PriceSum
            FROM samples."MS_SalesFacts_up" AS t1
            GROUP BY PaymentType, OrderYear, OrderMonth
            LIMIT 1000001`,
        },
    },
};

    salesSourceData: An arbitrary alias name you assign to the object with requested chart data, used for access on the Prepare tab.

    The sql_query field describes the query to the connection data source.

  4. Clear the contents of the Params and Config tabs: they contain a template that is not relevant to our example.

  5. On the Prepare tab, create a table:

    // Helper function to prepare source data in a more convenient format
function getPreparedLoadedData(loadedData, sourceName) {
    const sourceData = loadedData[sourceName];
    const columnNames = sourceData.filter(item => item.event === 'metadata').map(item => item.data.names)?.[0] || [];
    const res = [];
    sourceData.filter(item => item.event === 'row').map(item => item.data).forEach(rowItem => {
        const rowHash = rowItem.reduce((acc, field, index) => {
            acc[columnNames[index]] = field;
            return acc;
        }, {});
        res.push(rowHash);
    });
    return res;
}

// Helper function to group data by a specified field name 
function groupBy(arr, field) {
    return arr.reduce((acc, item) => {
        const key = item[field];
        if (!acc[key]) {
            acc[key] = [];
        }
        acc[key].push(item);
        
        return acc;
    }, {});
}

// Source data transformed into a convenient format
const data = getPreparedLoadedData(Editor.getLoadedData(), 'salesSourceData');

// Array containing unique values of the "OrderYear" field, sorted in ascending numerical order
const years = Array.from(new Set(data.map(d => String(d['OrderYear'])))).sort();

// Common styles for table header cells
const headStyles = {background: 'var(--g-color-base-misc-light)', verticalAlign: 'middle'};

// Table header cell configuration
const head = [
    {
        name: 'Payment type',
        formattedName: Editor.generateHtml({
            tag: 'span',
            content: [
                {tag: 'span', content: 'Payment type'},
				// tooltip for a cell header
                {
                    tag: 'dl-tooltip',
                    content: ' ℹ',
                    style: {
                        display: 'inline-block',
                        margin: '0px 0px 0px 4px',
                        'line-height': '12px',
                        'text-align': 'center',
                        width: '16px',
                        height: '16px',
                        border: '1px solid #ccc',
                        'border-radius': '50%',
                    },
                    attributes : {
                        'data-tooltip-content': {
                            tag: 'i',
                            content: 'Tooltip content',
                        },
                    },
                }
            ],
        }),
        css: headStyles,
        pinned: true,
    },
    // Creating columns based on the array of values from the "OrderYear" field obtained earlier
    ...years.map(year => ({
        name: year,
        css: headStyles,
    })),
    {
        name: 'Sales, all years',
        css: headStyles,
    },
];

// Helper function to render a chart line
function createChart(chartData) {
    const chartWidth = 80;
    const chartHeight = 40;

    // Calculating the minimum and maximum coordinate values
    const minX = Math.min(...chartData.map(d => d.x));
    const maxX = Math.max(...chartData.map(d => d.x));
    const minY = Math.min(...chartData.map(d => d.y));
    const maxY = Math.max(...chartData.map(d => d.y));

    // Calculating coordinates based on the chart container dimensions (chartWidth, chartHeight)
    const coords = chartData.sort((d1, d2) => d1.x - d2.x).map(d => ([
        (d.x - minX) / (maxX - minX) * chartWidth, 
        (d.y - minY) / (maxY - minY) * chartHeight,
    ]));
    // Creating a path for the SVG line using the coordinates generated above
    let d = "";
    coords.forEach((_, x) => {
        d += d === "" ? "M" : " L";
        d += `${coords[x][0]} ${coords[x][1]}`;
    });
    // Creating an SVG with var(--g-color-base-brand) for line color and thickness of 2px
    return `
        <svg width="${chartWidth}" height="${chartHeight}">
            <path 
                d="${d}" 
                style="fill: none; stroke: var(--g-color-base-brand); stroke-width: 2;"
            />
        </svg>`;
}

const rows = [];

// Helper function for number formatting
const formatSalesValue = new Intl.NumberFormat('ru-RU').format;
const postfix = ', ₽';

// Rows grouped by the "PaymentType" field
const groupedData = groupBy(data, 'PaymentType');

// Generating and populating table rows for each grouped payment type (based on the PaymentType field value)
Object.entries(groupedData).forEach(([key, items]) => {
    // Rows grouped by the "OrderYear" field
    const salesByYears = groupBy(items, 'OrderYear');
    // Calculating the sum for the "PriceSum" field across all years (based on OrderYear)
    const totalSales = items.reduce((sum, d) => sum + d['PriceSum'], 0);
    rows.push({
        cells: [
            {
                value: key,
            },
            // Creating columns based on previously prepared "OrderYear" values
            ...years.map(year => {
                const salesByYear = salesByYears[year] ?? [];
                const yearSales = salesByYear.map(d => ({
                    x: new Date(d['OrderMonth']).getTime(), 
                    y: d['PriceSum'],
                }));
                const maxSales = Math.max(...salesByYear.map(d => d['PriceSum']));
                const minSales = Math.min(...salesByYear.map(d => d['PriceSum']));

                return {
                    value: maxSales, 
                    formattedValue: Editor.generateHtml(`
                        <div>
                            ${createChart(yearSales)}
                            <div style="margin-top: 8px;">Min: <b>${formatSalesValue(minSales)}${postfix}<b></div>
                            <div>Max: <b>${formatSalesValue(maxSales)}${postfix}</b></div>
                        </div>
                    `),   
                };
            }),
            {
                value: totalSales,
                formattedValue: formatSalesValue(totalSales) + postfix,
                css: {
                    verticalAlign: 'middle',
                    textAlign: 'center',
                    fontSize: '16px',
                },
            },
        ],
    });
});

module.exports = {head, rows};

  6. At the top of the chart, click Execute. The preview will show the connection as a table with rows grouped by the PaymentType field and columns grouped by the OrderYear field, along with a monthly sales chart:

    image.png

  7. To save a chart, click Save in the top-right corner and enter a name for the chart.

Previous
Creating a table based on a dataset
Next
Creating a table via API