In this blog, we are going to review the Copy Data activity. The Copy Data activity can be used to copy data among data stores located on-premises and in the cloud.

In part four of my Azure Data Factory series, I showed you how you could use the If Condition activity to compare the output parameters from two separate activities. We determined that if the last modified date of the file (located in blob storage) was greater than the last load time then the file should be processed. Now we are going to proceed with that processing using the Copy activity. If you are new to this series you can view the previous posts here:

• Check out part one here: Azure Data Factory – Get Metadata Activity
• Check out part two here: Azure Data Factory – Stored Procedure Activity
• Check out part three here: Azure Data Factory – Lookup Activity
• Check out part four here: Azure Data Factory – If Condition Activity

If Condition Activity – If True activities

As mentioned in the intro, we are going to pick up where we left on in the last blog post. Remember that the If Condition activity evaluates an expression and returns either true or false. If the result is true then one set of activities are executed and if the result is false then another set of activities will be executed. In this blog post we are going to focus on the If True activities.

• First, select the If Condition activity and then click on activities from the properties pane, highlighted in yellow in the image below.
• Next, click on the box “Add if True Activity”.

This will launch a new designer window, and this is where we will be configuring our Copy Data activity.

Copy Data Activity, setup and configuration

Fortunately the Copy Data activity is rather simple to set up and configure, this is a good thing since it is likely you will be using this activity quite often. The first thing we need to do is add the activity to our pipeline. Select the Copy Data activity from the Data Transformation category and add it to the pipeline.

Now we need to set up the source and the sink datasets, and then map those datasets. You can think of the sink dataset as the destination for all intents and purposes.

Select the copy data activity and then click on the Source tab found in the properties window. For the dataset I am simply going to choose the same dataset we used in our first blog in this series.

Creating the Sink dataset in Azure Data Factory

Now it’s time to set up the Sink (destination) dataset. My destination dataset in this scenario is going to be an Azure SQL Database. I am going to click on Sink and then I will click on + New.

When you click on + New, this will launch a new window with all your data store options. I am going to choose Azure SQL Database.

This will launch a new tab in a Azure Data Factory where you can now configure your dataset. In the properties window select “Connection”, here you will select your linked service, this is your connection to your database. If you haven’t already created a linked service to your database then click on + New. Clicking new will open up a new window where you can input the properties for your linked service. Take a look at the screenshot below:

Once the linked service account has been created I will select the table I want to load. All available tables will show up in the drop down list. I will select my dbo.emp table from the drop down.

The final step for the dataset is to import the schema, this will be used when we perform the mapping between the source and sink datasets. Next, I am going to click on the Schema tab. On the Schema tab I will select import schema, this actually returns a column that I don’t need and so I will remove the column from the schema as well. Take a look at steps three and four in the following screenshot:

Copy Data activity – Mapping

Now that the source and sink datasets have been configured, it’s time to finish configuring the Copy Data activity. The final step is to map these two datasets. Now I will navigate back to the pipeline and click on the copy data activity. From the properties window I will select the Mapping tab. Then I will import the schemas, importing schemas will import the schemas from the source and the destination. This step will also attempt to automatically map the source and the destination as well and so it is important to verify that the mapping is correct. Unfortunately my source dataset was a file and it did not have any column headers. Azure Data Factory automatically created the column headers Prop_0 and Prop_1 for my first and last name columns. Take a look at the following screenshot:

This was a simple application of the Copy Data activity, in a future blog post I will show you how to parameterize the datasets to make this process dynamic. Thanks for reading my blog!

One of my favorite tricks in Power BI is to use R integration to quickly export data outside of Power BI. This is also a very popular question among Power BI enthusiasts. Once users realize the true capabilities and easy of the Power Query editor to transform and clean data they want to clean up their files using Power BI but the challenge is then how do I get the data out?

Update: Video embedded at bottom of blog post.

Prerequisite – R

If you want to test this method out you need to install R on your machine, for your convenience see the following link:

https://mran.microsoft.com/open/

Export data from Power BI using R

Like all of my quick tip blogs, this will be quick .

You will export data from the Power Query Editor. To launch the Power Query Editor click Edit Queries from the Home ribbon in Power BI.

Inside the Power Query Editor, click the transform ribbon and then click on R.

Once you click on the large letter R seen in the screenshot below a new window will open. Now you just have to type in some basic code. Here is the pseudo code:

write.csv(dataset, <destination>)

That’s it! Click ok and check your folder for the file. Do notice that I used two backslash characters, this is required. The other question I get is can I write the results to an excel file or to a SQL Database table and the answer is…. Yes! R has packages that make doing particular tasks easier and if you want to write to an Excel file or SQL Server table you would need to install those packages. My hope is to do some follow-up post around this topic.

Thanks for checking out my blog!

Thanks for watching my latest webinar on DAX. If you missed the webinar you can watch the recording here:

http://pragmaticworks.com/Training/Details/Advanced-DAX-Problem-Design-Solution

I didn’t have a lot of questions from this webinar but I will answer a the ones I did have.

Question: What is the difference between the Find function and the Search function.

Answer: The FIND function is case-sensitive and the search function is case-insensitive. In the webinar I used the FIND function in my first demo but really this could have been replaced with the SEARCH function since I was not concerned with case sensitivity.

Question: Where did you find the School Grade demo that you used in your webinar?

Answer: The data I used in the webinar can be found at the link below:

http://schoolgrades.fldoe.org/reports/#2014

Question: Might there be an upcoming webinar regarding DAX as it pertains to SSAS cubes rather than Power BI?

Answer: I am not aware of any webinars upcoming that will be specific to Tabular Cubes. It is important to point out that the examples used in this webinar are also applicable with Tabular SSAS Cubes since it uses the same backend engine as Power BI Desktop (xVelocity). If you implement these solutions in SSAS Tabular, the consumption of the data would be achieved through excel and then the necessary filters and slicers would be applied there instead of the report view provided in Power BI Desktop.

Thanks again!

Hey Everyone,

Thanks for attending the free 1 hour DAX Webinar by Pragmatic Works. In this blog I am going to focus on the primary question that I saw come up over and over again in the chat logs. The webinar can be found here:

• How come the totals for Forecast and Dynamic are the same?
• The Grand Total is different as well.
• Are we sure all those totals are right? for The MLS data?

The totals that are generated in DAX are usually an area that cause confusion, especially if you are still learning about filter context. Totals do NOT simply add up the values that appear in the rows on the report. This is confusing because most of the time it looks like this is exactly what is occurring in the background.

If the Totals cell in your table does not sum the rows then how does it come up with a total?? Great question! The total cell also executes the DAX formula defined in your measure. Let’s start with a visual, one from the webinar that prompted the questions leading to this post. Notice in this picture you can clearly see that the “TOTAL” is not the sum of the rows. The values represented in the total row look completely wrong, however they are not.

Remember this question? How come the totals for Forecast and Dynamic are the same? Let’s try to answer this question now. We can begin with Forecast YTD Sales.

Forecast YTD Sales is a simple calculation in the model. It is simply Prior YTD Sales * 2. In the screenshot below I have added the calculated measure Prior YTD Sales to the table so we can visualize exactly how the total cell is being calculated.

The total value is $19,582,120 which is $9,791,060 * 2.

Remember that in our original table there was only data up until June of 2008. Therefore, why is the total row taking Prior YTD Sales from December? Why is the total calculation taking the value of $9,791,060 from the Prior YTD Sales measure instead of $3,037,501 (June)? The table visual only displayed data up until June but this was because we don’t have any sales past June of 2008. The total row is looking at the entire 2008 year and therefore gets the Prior YTD Sales value as of the last date in 2008 ($9.791,060). Although this may not be the desired behavior it is the correct behavior.

How come the totals for Forecast and Dynamic are the same?

Well all of that is great, but that still doesn’t explain why the Forecast YTD Sales has the same value as our Dynamic Measure at the total row. The Dynamic Measure calculation always takes YTD Sales unless there are no sales on the last day of the current month in which case it will take the Forecast YTD Sales. This was explained in the webinar so please revisit that if you don’t remember the why behind this calculation.

IF(
    LastSaleDate <> LastDayOfMonth,
    [Forecast YTD Sales],
    [YTD Sales])

At the total row the calculation would read like this:

IF(
    “6/20/2008” <> “12/31/2008”,
    [Forecast YTD Sales],
    [YTD Sales])

Based on this interpretation of the measure we can see that the last day a sale took place in our model does not equal the last date in our model for the current year. That means that the total ROW is going to display the measure [Forecast YTD Sales]. This explains why we see the following values in our total row:

Once again understanding the total row is not easy, especially if you are transitioning to DAX from working with excel. If the total row seems incorrect it’s important to validate the calculated measure at the total level.

Thanks!