Power Query date format (How to + 5 tricky scenarios)

One thing that continually causes problems for all data professionals is dates. Over time, Microsoft has given us a plethora of tools and Excel functions for handling dates, yet they still catch us out. Power Query is a tool for manipulating and cleaning data, and it has a lot of features for managing dates. The Power Query date format is flexible and can be converted into many other text and numeric formats.

For this post, let’s begin by looking at the basics of Power Query date formats. Then once we’ve grasped that, we can move on to look at some tricky date formats scenarios.

Watch the video:

Watch the video on YouTube

Download the example file: Click the link below to download the example file used for this post:

Simple Power Query date format transformations

In this section, we look at the basic process and simple transformations for working with dates.

Date data type

When we get our data into Power Query, we first need to ensure date columns have the correct data type. Sometimes Power Query’s automatic detection of data types gets it wrong, so we need to check that all columns are correctly formatted.

To change a specific column into a date format, we have lots of options:

• Click the data type icon in the column header and select Date
• Select the column, then click Transform > Data Type > Date from the Ribbon
• Right-click on the column header, then click Change Type > Date
• Change the applied data type in the M code to type date

Changing the data type is the most important step. Until the columns are recognized as dates, Power Query cannot use them as dates.

Extracting additional information from a date

From a date column, we can extract other information. The Power Query date formats available from the ribbon are shown below.

These are available in Power Query’s Transform and Add Column Ribbons. The Transform ribbon converts the selected column while Add Column creates a new column based on the selection.

The following are examples of the types of information we can extract from the date column (Year, Days in Month, Week of Year, Day Name, Day of Year)

M code for date transformations

Each of the transformations above apply a Power Query function to extract the information from the date. For example, the M code to add the Year and Day of Year columns are.

Year

= Table.AddColumn(#"Changed Type", "Year", each Date.Year([#"Date (YYYY-MM-DD)"]), Int64.Type)

The Date.Year function extracts the Year from a date.

Day of Year

= Table.AddColumn(#"Inserted Day Name", "Day of Year", each Date.DayOfYear([#"Date (YYYY-MM-DD)"]), Int64.Type)

The Date.DayOfYear function calculates the year day number from a date.

Additional information:

• Find out more about Power Query functions here: Power Query formulas (how to use them and pitfalls to avoid).
• Details of all the date functions are here: https://learn.microsoft.com/en-us/powerquery-m/date-functions

Tricky transformations

The biggest issue we face is trying to change a column into a date type initially. Often this occurs when working with different country date formats.

The date format used in the US is month – day – year. Yet, it is day – month – year in the UK. Then there is the ISO 8601 format, which is year – month – day. 

Personally, I prefer the year – month – day format as it can be sorted correctly even when it is text or numeric. But I can’t imagine the UK government is about to change anything on my account.

Sometimes, Power Query can get confused if we use data containing dates in a format different from our local settings. For example, if the cell value is 01/02/03, it could be:

• 2nd January 2003 – the US format
• 1st February 2003 – the UK format
• 3rd February 2001 – the ISO 8601 format

In this section, we look at how we force Power Query to understand dates, even if different country date formats are used.

Examples

Look at the screenshot below. It shows dates in five different formats (without any data types applied). I have included the format in the header so you can see what the date format is meant to be.

NOTE: I am based in the UK. My regional settings are set to English (United Kingdom), so the dd.mm.yyyy is the expected format for dates. Depending on your location settings, you may get different results to those shown below.

Using Power Query, I translate each of the columns into a date data type.

• dd.mm.yyyy – as expected, Power Query is happy to translate a date in my local region format.
• mm.dd.yyyy – the first row is an error because the value 07.23.2017 as a UK date would be the 7th day of the 23rd month. This is not a valid date in my region. In the second row, the value is 09/07/2017. As a UK format, this would be the 9th day of the 7th month, which is a valid date. However, even though the format is valid, the result is incorrect. The original date is in a mm.dd.yyyy format, so the correct value would be 7th day of the 9th month.
• yyyy.mm.dd – due to the year having a 4 character format, Power Query has correctly guessed that the remaining digits are month followed by day.
• yy.mm.dd – the data for this example includes the year with just two digits. Power Query can’t work out which digits relate to the year, month or day. It has applied my local format. However, this has mixed up the day and year, meaning every value is incorrect. This is the most dangerous result, as if we don’t notice and keep going, it could cause havoc later on.
• yyyymmdd – there are no separators in this example. Power Query has not recognized this as any date, so each value displays as an error.

Power Query gives us lots of tools to solve each of these date formats.

Locale settings

If the date is a valid country format, we can explicitly tell Power Query which country to use.

Click the data type icon and select Using Locale… from the menu.

From the Change Type with Locale dialog box, select the following options:

• Data Type: Date
• Locale: The region for the date format (for mm.dd.yyyy, I’ve selected English (United States))

Then click OK to close the window.

Power Query now shows the correct date.

In the previous section, Power Query already got the correct result for the yyyy.mm.dd format, so we can move on to the next column.

To change the yy.mm.dd format, we can select English (Canada) as the locale date format. At the bottom of the window, it displays a sample of formats that are correctly recognized.

The yyyy-mm-dd format is the first item in the list of sample inputs (shown in the purple box above). This order matches with our example data. Selecting this option generates the correct values for this column. Power Query is intelligent enough to recognize the separators and create the correct date.

It may take a bit of exploration to find out which locale (if any) match the format we’ve got.

While the yyyymmdd format is in the same order as English (Canada), due to the lack of separators, Power Query doesn’t recognize this as a date. We will come back to this example in a moment.

Changing the workbook Locale setting

If a workbook is going to have a lot of dates in a specific region format, we can change the Locale settings for the whole workbook.

Click File > Options and settings > Query Options

In the Query Options dialog box, select Regional Settings > [Required Locale] > OK

This setting applies to the current workbook only.

Once the locale is set, we only need to use the standard date data type and Power Query will apply that locale setting to each field.

Dealing with unknown date formats

We were left with the tricky yyyymmdd format. There are a few options we can try here. The two we are going to look at are:

• Power Query Formulas
• Column From Examples

Using Power Query formulas

Our date is in a standard format, so we can apply basic M code formulas to calculate the result.

Click Add Column > Custom Column

In the Custom Column dialog box, I have entered the following formula.

=Text.End(Text.From([yyyymmdd]),2) & "/" 
    & Text.Middle(Text.From([yyyymmdd]),4,2) 
    & "/" & Text.Start(Text.From([yyyymmdd]),4)

This is constructed using 4 functions:

• Text.From() – converts the value in the field to text
• Text.Start() – similar to Excel’s LEFT function to return the first characters from a text string
• Text.End() – similar to Excel’s RIGHT function to return the last characters from a text string
• Text.Middle() – similar to Excel’s MID function to extract characters from the middle of a text string

With this formula, we’ve constructed a text string that generates the date in a UK format. Power Query can now easily recognize this as a date.

This formula is specific for this scenario. Change the formula to match your requirements.

Using Column From Examples

Column From Examples is a feature that helps to standardize formats with complex structures.

Select the column, then click Add Column > Column From Examples (drop-down) > From Selection

In the column on the far right, we can start typing the date we want to see, then press Enter. For the value 20170723, I have entered 23/07/2017 as shown by the screenshot below.

Power Query does its magic and creates values for all our rows. Check that these are right. If not, correct any that are wrong; each time we provide more detail, Power Query adjusts its suggested transformation.

This is an advanced tool. If there isn’t enough standardization in the value, Power Query can easily calculate the incorrect results. But for our scenario, it is reasonably straightforward.

Once happy with the format, click OK.

Conclusion

We have seen how to use the Power Query date format to extract additional information from a date. These are quick and easy transformations. However, before we are able to make those transformations, we need the column to be a date data type.

Using different date formats, we were able to get standardized dates from all of them. Therefore, no matter where the data comes from, we can convert it and use it in our local region.