Power BI

Select rows that have no empty fields using Expression.Evaluate in Power BI and Power Query

by

This article describes the use of Expression.Evaluate as a very helpful “swiss-army-knife”-method for your Power BI toolbox as it has many more use cases than the one described below. It lets you perform repeating tasks without using functions and can even replace recursive operations in some cases – but that’s a topic for a later blogpost. Today we start with a simple task:

If you want to filter rows only where none of the fields/columns are empty, you’d either pass not-null filters into every single column (which is tedious and not dynamic), transpose or add an index column, unpivot and filter from there (both might drag performance down on large tables).

Using Expression.Evaluate will create a dynamic approach that doesn’t force you to transform your source table nor to directly specify your column names: It will take in a text string and execute it as if you’ve been written it manually into the formula. If you manage to let this textstring be created dynamically according to the different tables you’re going to pass in – Expression.Evaluate will make sure that these individual statements will be executed accordingly.

So for a table containing Column “A” and Column “B” the filter statement would look like this:

Read more

How to create a Load History or Load Log in Power Query or Power BI

by

If you want to collect your regularly loaded data without overwriting it or create a load log that writes the load activites with a timestamp into a table, you need to create a query that adds new lines to it’s own latest version.

In both cases you add a column that returns the current day and time using DateTime.LocalNow()

Image1

Load Log the PowerBI-way

Then using Power BI it’s fairly easy: Use the R-extension to perform an append query to the external txt or csv-file like described in this post (incremental load):

require(gdata)
write.table(trim(dataset), file=”your filepath & filename.txt”, sep = “\t”, row.names = FALSE, append = TRUE, column.names=FALSE)
plot(dataset);

Load Log the Power Query-way

In Excel’s Power Query it’s a bit trickier, because you cannot export your queries and therefore need to create the consolidated table in your workbook itself. And it must be a self-referencing query or to be more precise: a self-sourcing query: A query that takes its latest output as its new input, then appends the new data and overwrites the old result with the new one.

How do we deal with this? When we write the query we don’t have the result yet to be referenced?

Read more

Trap in Replacement function in Power Query/M/PowerBI

by

Recently Power Query’s replacement function greeted me with some unexpected results: I applied the same steps but sometimes replacements of blanks with nothing did what it should and sometimes not.

What was going on?:

Starting with a table with 2 columns like this:

Image1

The results came out as expected, blanks were removed by nothing, so all gaps closed:

Image2

Then I added the Values-column like this and applied the same steps as above:

Image3

 

Image4

Read more

Visualize query dependencies in Power BI with Sankey diagram

by

Edit 7th August 2016: You’ll find an easier “OneInAll”-approach in the PowerBI forum here. Thanks to David Moss for reporting an issue with the solution below.

PBI-file with new query: NewAllInOneCode.zip ,

How to use:

Old post:

This week we got a new Sankey diagram. It has some advantages when visualizing dependencies between items/nodes, one being not to cut the labels. Time to share a tool that I use a lot when working with complex M-queries (in Power Query or Power BI). It shows how the queries are connected with each other:

Image1
Sankey diagram default

Left stand the queries who fetch the source data, parameters and/or user defined filters, then you see how they get transformed until to the right you see the queries that deliver your end-results. Especially when working with codes that has been written by someone else I find it very useful to get a quick oversight of what’s happening – as an excellent complement to the documentations 🙂

You can also move the nodes around to improve readability:

Image2
Sankey diagram with adjusted nodes

Read more

Use R to export data from Power BI

by

Edit 03 Aug 2016: With the July 2016 release you can now run your R-scripts directly in the query-editor (so no row-limit any more!). No need to go via the visuals. But this will limit you to export datasets from the query-editor, so no DAX.

Edit 22 Jan 2016: Currently this method is limited to 150k rows! It will drop all others without warning!

Edit 25 Jan 2016: 150k rows is the limit on all R-scripts at the moment (link). A warning sign will be shown on the image. Will create a new blogpost if this changes – so stay tuned (by following this blog, Twitter or LinkedIn)

With the December release, Microsoft enabled Power BI’s content to be fetched from the R-feature. So instead of plotting this data to the screen, could we just export it to a file?

Yes we can:

write.table(trim(dataset), file=”your filepath & filename.txt”, sep = “\t”, row.names = FALSE);

write.table has a lot of options, which are explained here. I’ve only used a few here: Exactly defining the location where the output-file shall be stored (file=), setting tab as delimiter (sep=”\t”) and skipping row names, which would be an index in our example here.

In addition to that, I had to get rid of trailing whitespaces that somehow sneaked into the table (trim(dataset)). Therefore the gdata-package is needed. So the full R-code looks like this:

require(gdata)
write.table(trim(dataset), file=”your filepath & filename.txt”, sep = “\t”, row.names = FALSE)
plot(dataset);

Here you have to make sure to turn the slashes in the filepath: So if your path looks like this: C:\Users\Imke\Desktop\Rfile.txt you need to write it like this: C:/Users/Imke/Desktop/Rfile.txt

NEW: You find a function for this export here: https://github.com/ImkeF/M/blob/master/LibraryR/Table.ExportToCsv.pq , it swaps the slashed automatically.

Let’s take this as a start that will probably fulfill many users need to quickly export data from Power BI to simple text files. Further use cases I see are:

Read more

Visualize dependencies between your DAX measures: DAX-VizArt-Wizard

by

This DAX-VizArt-Wizard vizualizes dependencies between your DAX measures, shows the definition of all related measures and shows differences between the measures of 2 models/versions. This works for Power Pivot, Power BI and for Analysis Services Tabular (SSAS).

In the Power BI-Version you’ll see them in the Sankey-chart like this:

VizArt1

If select measures/nodes, all direct connections will be highlighted:

VizArt2Direct

In the second version, all indirect connections will be highlighted as well & the selected measure definitions will be shown.

Read more