Category: Microsoft

Posted on

Power BI – Quick Measures

Microsoft has prepared Power BI to make it easier for users to perform DAX measures without a lot of DAX knowledge.

To access the Quick Measures, click QUICK MEASURES on the Home tab. You will see a list of options to choose from. On the right side of the Quick Measure dialog box, you will find all the tables from your data model.

DAX

Example 1 – Rolling Average

Many use rolling average to smoothing the data set. Unusual periods can be disrupting for understanding patterns and especially in projections, they can make forecasts unnecessarily inaccurate.

In the example below a line chart visualises sales numbers over several years, but a number of periods were unusual and you want the audience to understand how the quantity would look under normal conditions.

DAX

From the dialog box you can select the Quick Measure ‘Rolling Average’ from the Calculation list. You will now need to add the fields from your tables and set up the parameters.

In this example the quantity is the Base value, Dates are added to the Date field (Rolling Average is a Time Intelligent measure and the primary key from the timetable needs to be added), and Periods before and after are set to 3 months (the smoothing level).

Measures

When you click OK Power BI will the write the DAX.

Quantity rolling average  =

IF(

ISFILTERED(‘Dates'[Dates]),

ERROR(“Time intelligence quick measures can only be grouped or filtered by the Power BI-provided date hierarchy or primary date column.”),

VAR __LAST_DATE = ENDOFMONTH(‘Dates'[Dates].[Date])

VAR __DATE_PERIOD =

DATESBETWEEN(

‘Dates'[Dates].[Date],

STARTOFMONTH(DATEADD(__LAST_DATE, -3, MONTH)),

ENDOFMONTH(DATEADD(__LAST_DATE, 3, MONTH))

)

RETURN

AVERAGEX(

CALCULATETABLE(

SUMMARIZE(

VALUES(‘Dates’),

‘Dates'[Dates].[Year],

‘Dates'[Dates].[QuarterNo],

‘Dates'[Dates].[Quarter],

‘Dates'[Dates].[MonthNo],

‘Dates'[Dates].[Month]

),

__DATE_PERIOD

),

CALCULATE(SUM(‘Line_Items'[Quantity]), ALL(‘Dates'[Dates].[Day]))

)

)

 

By adding the Rolling Average to the line chart, you can see the result below.

Measures

Example 2 – Percentage difference from filtered value.

The line chart below (the columns) shows sales in 3 different countries Canada, Mexico, and United States. It shows how much is generated in sales by Canada and Mexico by percentage.

The line is calculated by another Quick Measure – Percentage difference from filtered value.

Measures

In this example the Base Value is a sales measure. You can define how you want this quick measure the handle blanks. You can display them as blanks or you can tell the measure to treat blanks as zero. In this example the measure is filtered by country, and we have selected United States.

Measures

And as in the first example Power BI will write the DAX.

sales % difference from U.S.A. =

VAR __BASELINE_VALUE = CALCULATE([sales], ‘Customers'[Country] IN { “U.S.A.” })

VAR __MEASURE_VALUE = [sales]

RETURN

IF(

NOT ISBLANK(__MEASURE_VALUE),

DIVIDE(__MEASURE_VALUE – __BASELINE_VALUE, __BASELINE_VALUE)

)

Example 3 – Correlation Coefficient

In this example correlation between product unit prices and sales quantity needs to be investigated. Does the price affect the quantity sold?

The quantity and unit price have been added to a scatter chard below. The trend line will indicate to us the relationship between the two. The dots are spread out. This indicates that there isn’t a close relationship between the two variables, but what is the correlation coefficient?

A Quick Measure can very simply find the result.

Category here is the products identified by the field Product ID. Measure X and Measure Y are the here the sum of quantity and sum of unit price.

Again, Power BI will write the DAX.

Sum of Unit Price and Quantity correlation for Product ID =

VAR __CORRELATION_TABLE = VALUES(‘Items'[Product ID])

VAR __COUNT =

COUNTX(

KEEPFILTERS(__CORRELATION_TABLE),

CALCULATE(SUM(‘Items'[Unit Price]) * SUM(‘Line_Items'[Quantity]))

)

VAR __SUM_X =

SUMX(

KEEPFILTERS(__CORRELATION_TABLE),

CALCULATE(SUM(‘Items'[Unit Price]))

)

VAR __SUM_Y =

SUMX(

KEEPFILTERS(__CORRELATION_TABLE),

CALCULATE(SUM(‘Line_Items'[Quantity]))

)

VAR __SUM_XY =

SUMX(

KEEPFILTERS(__CORRELATION_TABLE),

CALCULATE(SUM(‘Items'[Unit Price]) * SUM(‘Line_Items'[Quantity]) * 1.)

)

VAR __SUM_X2 =

SUMX(

KEEPFILTERS(__CORRELATION_TABLE),

CALCULATE(SUM(‘Items'[Unit Price]) ^ 2)

)

VAR __SUM_Y2 =

SUMX(

KEEPFILTERS(__CORRELATION_TABLE),

CALCULATE(SUM(‘Line_Items'[Quantity]) ^ 2)

)

RETURN

DIVIDE(

__COUNT * __SUM_XY – __SUM_X * __SUM_Y * 1.,

SQRT(

(__COUNT * __SUM_X2 – __SUM_X ^ 2)

* (__COUNT * __SUM_Y2 – __SUM_Y ^ 2)

)

)

 

The correlation between unit price and quantity here is -0.15. In other words when a product gets more expensive the sold quantity decreases. But a negative correlation of -0.15 isn’t much. Nevertheless it could be a clever idea to keep an eye on this number over time to understand the sales pattern.

Conclusion

Microsoft has given their clients a shortcut to create DAX measures by offer the quick measure tool. Quite complicated measures can be achieved without prior DAX knowledge.

Posted on

Tracking Performance using Conditional Formatting in Excel

Conditional Formatting is one of the most powerful tools in Excel. It can do so much to help you track data with relatively little time and effort. This ‘Quick Win’ tool can be mastered in a few simple steps. The blog will look at what Conditional Formatting is, how to apply it to data and what it can be useful for to improve performance tracking.

What is Conditional Formatting

Conditional Formatting is based on setting rules or conditions on specific data and if any of this data meets a rule then it will change its appearance in some way e.g. display an icon or change colour (see below)

conditional

 

In this example there are 3 separate rules all set according to specific bands or ranges of numbers. If a number falls into a specific band, then it will change to the appropriate colour. The world of finance has coined the term ‘RAG’ status – or Red, Amber, Green – to represent low, medium and high numbers. Excel has many pre-set icons and colour ranges to help you adopt this useful tracking system.

 

Why use Conditional Formatting

Think of situations where you might find this technique useful. For example, someone in HR might need to track all staff whose monthly sick days are greater than 4. Or an Events Organiser might want to colour code for each client on their books. Maybe an administrator needs to show when deadlines are about to be hit within 2 days, hit already or deadline has already passed all in different colours. Whatever the scenario, Conditional Formatting can vastly improve your data analysis techniques thus making you more productive.

 

How to apply Conditional Formatting

  1. Open ‘Revenue Table’ or similar dataset in Excel.
  2. Highlight all the sales figures and go to HOME > CONDITIONAL FORMATTING > HIGHLIGHT CELLS RULES:

formatting

  1. Select BETWEEN from the sub-list. This action brings up the following dialog box with red already set to a specific band:

  1. To customise your own bands, start with the highest first. Enter 100000 (lower limit) and 150000 (upper limit). From the ‘Drop down colour selector’, choose ‘Green Fill with Dark Green text’:

conditional

 

  1. With the figures still highlighted, repeat steps 2-4 setting the following rules/colours:

conditional

  1. The really amazing feature of Conditional Formatting once it’s created is that it is dynamic, i.e. the data will change colour if it meets the corresponding rule.
  2. For example, Ann Dodsworth’s revenue has increased to £51,000 so her revenue figure has changed from red to yellow (or amber):

Conclusion

Conditional Formatting in Excel can really help you understand your data in order to make insightful business decisions which in turn leads to greater profitability.