kb(pivotgrid): Add KB for local date aggregations and explain differences between local and XMLA binding

components/pivotgrid/data-binding.md

@@ -20,15 +20,22 @@ The PivotGrid supports different data sources via its `DataProviderType` paramet
 * [`Local`](#local) (default)
 * [`Xmla`](#xmla)
 
+### Usage Differences
+
+The available data providers differ in several ways:
+
+* XMLA binding is more complex to setup, but more flexible.
+* Local data does not support aggregations by date periods for `DateTime` properties. If a `DateTime` property is added as a row or column, the PivotGrid will generate a separate row or column for each unique `DateTime` value. If you need [aggregations by year, month, week, and so on, then create additional `int` or `string` properties in the PivotGrid model class](slug:pivotgrid-kb-local-date-aggregates).
+* XMLA binding supports [load on demand](slug:pivotgrid-overview#pivotgrid-parameters), which offloads all calculations to the external data source. Local binding receives all data at once and performs all aggregate calculations in-memory. Large amounts of local data may impact the performance, especially in WebAssembly apps.
+* When using load on demand, XMLA binding supports custom aggregate functions that are defined and performed in the OLAP cube. Local data supports only the [predefined aggregate types in the `PivotGridAggregateType` enum](slug:telerik.blazor.pivotgridaggregatetype).
+* When using local data, all defined measures in `<PivotGridMeasures>` render by default in the PivotGrid. Users can uncheck and hide the measures they don't need from the [PivotGrid Configurator](slug:pivotgrid-configurator).
 
 ## Local
 
-When bound to local data, the Pivot Grid requires its `Data` parameter to provide all the data at once as `IEnumerable<TItem>`. The component will perform all aggregate calculations in-memory and there is no [load on demand](slug:pivotgrid-overview#pivotgrid-parameters).
+When bound to local data, the Pivot Grid requires its `Data` parameter to provide all the data at once as `IEnumerable<TItem>`.
 
 If the local data changes programmatically, you need to reset the collection instance or [call the PivotGrid `Rebind()` method](slug:pivotgrid-overview#pivotgrid-reference-and-methods). See the common documentation about [refreshing component data](slug:common-features-data-binding-overview#refresh-data) for details.
 
-> Large amounts of local data may impact the performance, especially in WebAssembly applications.
-
 >caption PivotGrid bound to Local data provider
 
 <div class="skip-repl"></div>

knowledge-base/pivotgrid-local-date-aggregates.md

@@ -0,0 +1,169 @@
+---
+title: Use Date Aggregates with Local Data
+description: Learn how to implement and configure aggregate calculations by date periods like year, month, week, and day, when using local data binding with the Telerik PivotGrid for Blazor.
+type: how-to
+page_title: How to Use Date Aggregates with Local Data
+slug: pivotgrid-kb-local-date-aggregates
+tags: blazor, pivotgrid, aggregates
+ticketid: 1678109
+res_type: kb
+---
+
+## Environment
+
+<table>
+    <tbody>
+        <tr>
+            <td>Product</td>
+            <td>PivotGrid for Blazor</td>
+        </tr>
+    </tbody>
+</table>
+
+## Description
+
+This KB answers the following questions:
+
+* How to allow users to aggregate local data by date periods, such as years, months, weeks, and days?
+* How to enable date aggregations with local PivotGrid data binding?
+* How to calculate date aggregates by period with a local data provider?
+
+## Solution
+
+The [PivotGrid supports date aggregates by period](slug:pivotgrid-data-binding#usage-differences) only with [XMLA data binding](slug:pivotgrid-data-binding#xmla). Consider the following approach for local data scenarios:
+
+1. Define all `DateTime` properties in the PivotGrid model class as `internal`, `protected`, or `private`, so that the PivotGrid ignores them.
+1. Define additional helper properties in the PivotGrid model class, which will extract the year, month, week, and day values from the `DateTime` properties.
+1. Use the helper properties in the PivotGrid definition and UI.
+1. (optional) Use a [PivotGrid column header template](slug:pivotgrid-templates#column-header-template) to format the values of the helper properties. This may be necessary, because the PivotGrid sorts `string` values alphabetically in the column headers. In this case, you may also need to [get the correct localized value for the "TOTAL" label](slug:globalization-localization).
+
+>caption Using date period aggregations in the Telerik PivotGrid for Blazor
+
+````RAZOR.skip-repl
+@using System.Globalization
+@using Telerik.Blazor.Services
+
+@inject ITelerikStringLocalizer TelerikStringLocalizer
+
+<TelerikPivotGridContainer>
+
+    <TelerikPivotGridConfigurator />
+
+    <TelerikPivotGridConfiguratorButton />
+
+    <TelerikPivotGrid Data="@PivotData"
+                      DataProviderType="@PivotGridDataProviderType.Local"
+                      ColumnHeadersWidth="150px">
+        <PivotGridColumns>
+            <PivotGridColumn Name="@nameof(SalesModel.Year)" Title="Year" />
+            <PivotGridColumn Name="@nameof(SalesModel.Month)" Title="Month" />
+            <PivotGridColumn Name="@nameof(SalesModel.Week)" Title="Week" />
+        </PivotGridColumns>
+        <PivotGridRows>
+            <PivotGridRow Name="@nameof(SalesModel.Category)" Title="Category" />
+            <PivotGridRow Name="@nameof(SalesModel.Product)" />
+        </PivotGridRows>
+        <PivotGridMeasures>
+            <PivotGridMeasure Name="@nameof(SalesModel.Revenue)"
+                              Title="Revenue"
+                              Format="{0:C2}"
+                              Aggregate="@PivotGridAggregateType.Sum" />
+        </PivotGridMeasures>
+        <ColumnHeaderTemplate>
+            @{ var c = (PivotGridColumnHeaderTemplateContext)context; }
+            @ConvertYearMonthString(c.Text)
+        </ColumnHeaderTemplate>
+    </TelerikPivotGrid>
+
+</TelerikPivotGridContainer>
+
+@code {
+    private List<SalesModel> PivotData { get; set; } = new List<SalesModel>();
+
+    protected override void OnInitialized()
+    {
+        TotalSuffix = " " + TelerikStringLocalizer["PivotGrid_TotalValueFormat"].Replace("{0}", string.Empty).Trim();
+
+        GenerateData();
+    }
+
+    /// <summary>
+    /// Gets or sets the correct localized TOTAL string.
+    /// </summary>
+    private string TotalSuffix { get; set; } = " Total";
+
+    /// <summary>
+    /// Convert a <c>yyyy MM</c> string to <c>MMM yyyy</c>.
+    /// </summary>
+    /// <param name="yearMonth">The PivotGrid column header text that will render without a template.</param>
+    /// <returns>If the <c>yearMonth</c> argument is valid, returns a DateTime string in format <c>MMM yyyy</c> plus the optional Total suffix.
+    /// Otherwise, return the argument as is.</returns>
+    private string ConvertYearMonthString(string yearMonth)
+    {
+        string yearMonthWithoutTotal = yearMonth.Replace(TotalSuffix, string.Empty, StringComparison.InvariantCultureIgnoreCase);
+        string totalSuffixConditional = yearMonth.Replace(yearMonthWithoutTotal, string.Empty);
+        string[] separateYearMonth = yearMonthWithoutTotal.Split(" ");
+
+        if (yearMonthWithoutTotal.Length == 7 &&
+            yearMonthWithoutTotal.IndexOf(" ") == 4 &&
+            separateYearMonth.Length == 2)
+        {
+            bool yearSuccess = int.TryParse(separateYearMonth[0], NumberStyles.None, CultureInfo.InvariantCulture, out int year);
+            bool monthSuccess = int.TryParse(separateYearMonth[1], NumberStyles.None, CultureInfo.InvariantCulture, out int month);
+
+            if (yearSuccess && monthSuccess)
+            {
+                DateTime actualDate = new DateTime(year, month, 1);
+
+                return $"{actualDate.ToString("MMM yyyy")} {totalSuffixConditional}";
+            }
+        }
+
+        return yearMonth;
+    }
+
+    private void GenerateData()
+    {
+        int dataItemCount = 100;
+        int categoryCount = 3;
+        int productCount = 7 + 1; // effectively 7
+        int cityCount = 3 + 1; // effectively 3
+        int daysBack = 6 * 30;
+        Random rnd = Random.Shared;
+
+        for (int i = 1; i <= dataItemCount; i++)
+        {
+            var productNumber = rnd.Next(1, productCount);
+
+            PivotData.Add(new SalesModel()
+            {
+                Category = $"Category {productNumber % categoryCount + 1}",
+                Product = $"Product {productNumber}",
+                City = $"City {rnd.Next(1, cityCount)}",
+                ContractDate = DateTime.Today.AddDays(-rnd.Next(1, daysBack)),
+                Revenue = rnd.Next(123, 987) * 1.23m
+            });
+
+            PivotData.OrderBy(p => p.ContractDate);
+        }
+    }
+
+    public class SalesModel
+    {
+        public string Category { get; set; } = null!;
+        public string Product { get; set; } = null!;
+        public string City { get; set; } = null!;
+        internal DateTime ContractDate { get; set; }
+
+        public int Year => ContractDate.Year;
+        public string Month => ContractDate.ToString("yyyy MM");
+        public string Week => $"Week {ISOWeek.GetWeekOfYear(ContractDate)} {ContractDate.ToString("yyyy")}";
+
+        public decimal Revenue { get; set; }
+    }
+}
+````
+
+## See Also
+
+* [Pivot Grid Data Providers](slug:pivotgrid-data-binding)