Let's create a dataset and use this as the source.
We can now add some steps to the data flow to manipulate the data. We now spot that there is a new additional option, the 'Transform Column'.
In prior versions it was possible to carry out 'Transformations' but it required a few extra steps if we didn't want to retain the untransformed column. For example let us say that in the resulting data set we don't want to see the actual pupil numbers, we want to apply a rule to determine if the numbers are above or below a threshold and see that as a text description. Prior to having the Transform Column we could have created a new column in the dataflow, performed a function on that new column and then deleted the old column. Now we can use the Transform Column and do it all in a single step. We will apply the rule that if the school has over 200 pupils then it has 'A Lot' else it has 'A Few'. When I add the Tranform Column it pops up a window with a number of features, one part of which allows me to click on the step name and change that to anything I want. So below is the default name before I click on it ....
If I change this to say 'School Numbers' then we can see that the step changes in the flow, so rather than just showing the generic step type, we have an icon of the type of event that is being applied and a name depicting what is actually happening. This is useful to document the steps of the flow and make it easier to read.
I will now write the transformation statement as a simple CASE statement in the editor and then press Validate just to ensure that it is syntactically correct and then press Apply and we should see the column transformation take place.
Overall, quite a useful addition to the dataflow process. There's actually been quite a few nice little tweaks added to Dataflows and I encourage you to have a look at the release notes to see them. To finish let's have a look at a good aesthetic one. Now you'll see that the flow we have is very small and thus we can benefit from seeing all the descriptions.