Posts Tagged ‘Power BI’

Extract only letters from a mixed string in Power Query and Power BI

This is a quick method about extracting only letters from a string. It is part of the Week2 “Preppin’ data” challenge.

Task for extracting letters from a string

Image you have a string like so: “10.ROADBIKES.423/01” and would like to extract only “ROADBIKES”.
Power Query actually has a function for this purpose: Text.Select. It takes 2 parameters:

  1. The text to select from
  2. A list of characters that shall be selected

For the given example the code would look like so:

Text.Select( "10.ROADBIKES.423/01", {"A".."Z"})

This function is always case sensitive as there is no optional parameter that accepts a comparer function.

Easy application

Read more

Clean up or harmonize mis- or differently spelled category data with Power Query

A typical problem with data that has been created by manual entries is that category values are often misspelled or missed. So in this article I’m showing a very powerful technique on how to deal with this problem to clean up dirty category data. It was inspired by the “Preppin’ data” challenge whose instructions you can read here.

Task

Categorize dirty data

Read more

Your Oracle data import in Power BI and Power Query is slow?

If you’re using the native Oracle connector in Power Query, you will probably experience a very slow import performance. Thanks to Tristan Malherbe for recommending to use the OleDB-connector in Power Query instead. This speeds up import enormously.

How to create the connection string

If you’re using the OleDb.DataSource connector instead, you have to pass a connection string as the first parameter and an optional query record as the second parameter. To speed it up even more, you should use a FetchSize parameter in the connection string. For me, this didn’t work when I pasted it into the popup-window. So I had to manually add it in the query editor:

OleDB connection to an Oracle database

The “:1521” in the connection string is the port number, which is usually 1521 for Oracle databases.

Read more

Convert DateTime to ISO 8601 date and time strings in Power Query

Often, when querying APIs it is required to enter date and time filters in ISO 8601 format . Today I show a quick way to convert DateTime to ISO 8601 string, based on an ordinary DateTime field according to the following pattern:

2020-10-11T15:00:00-01:00

This represents the 11th October 3pm in UTC -1 timeszone.

Steps to convert DateTime to ISO 8601

If I enter:

#datetime(2020,10,11,12,0,0)

into the formula bar, it will be converted to :

11/10/2020 12:00:00

Comparing to the desired ISO format the year, month and days are in the wrong order. So using the universal Text.From function will not return the correct result.

Read more

Retrieve header fields like response status from Web.Contents in Power BI and Power Query

Many Power Query function not only return their values as advertised in their function documentation, but on top of that a metadata record. This record is like tag that holds additional information about the returned main value (for more details about this, please check out my friend Lars Schreiber’s article about it).

Useful metadata for the Web.Contents function

Today I discovered that the function Web.Contents delivers a really nice record with a couple of useful information. To retrieve header fields, you have to use the Value.Metadata function, like so for example:

Return header fields like response status from Web.Contents

Interesting metadata from the Web.Contents – function

This might help for some advanced web query tasks.

How to use

If you want to use this in production, you’d probably branch out the logic. So first use Web.Contents and keep that result in a column or variable. Then add another column that references it and return the metadata record.
Apply the logic check on it and create a last column where you finally parse the content from the binary that Web.Content has returned.

Enjoy & stay queryious 😉

Transform a query into a function in Power Query and Power BI

In my previous blogpost I’ve described a method how to extract a substring that follows a certain pattern from a string. In this post I show how to transform a query into a function that can be applied to many rows of a table.

Video how to transform a query into a function

Please check the video for detailed steps. In there I also show how to modify the code. It shall also detect strings with a sequence of just 8 numbers. In the original query, those had to be followed by a minus sign and another number:

Read more