You can download the file from here.

Share your feedback's and comment's below

As all of you might know,I am an Analytics Guy and always intrigued by numbers. This led me to understand my connections in LinkedIn. In the process I saw an graph which was absolutely stunning and tickled my brains to do some work to replicate the same in SQL Server Reporting Services. I always believed and still believe that SQL Server can do anything! absolutely anything. It’s just that you know how to use it.

Below was the chart which made me wear my thinking hat!

So, what’s so intriguing about it? Yeah, it looks more like a scatter plot chart with varying size depending on the measure? hmmmmm…….

Here is the catch, You need to have the largest circle(Of course it has highest radius!) in the middle and rest of the circles surrounding it with appropriate radius. Here, I have the highest connection with a company X which has around 77 connections in the middle.

Well, Its not that tough when you know SQL Server spatial will solve your problem to create such a intriguing chart.

Enough of background! So, How was it done?

• Created a function which will return the point co-ordinates of each location within equal interval depending on the number of companies.
• Created a function which will return the radius of each circle depending on the measure size.
• Created a SSRS report to do the final touches

Here is the output in SQL Server and after little bit of tinkering the output in SSRS :

I have uploaded the scripts and Reports in my skydrive. Feel free to download and showcase this to your client and partners.

Happy reading and enjoy the weekend.

Have you anytime been knocked on the head by the client on how can you as a data modeler taken actions to reduce space usage in the server?

One of the most oblivious to any data/dimensional modeler is the efficient way of minimizing the space used by the database on the disk. So how can we  achieve that? Below is one of the trick which can be used. Sparse columns in SQL Server. It is an fun-tas-tic piece of thing which can save you lot of space of your database.

1. What is a sparse column?

A sparse column in  a table is the one which contains a lot of NULL values. It may be required as per design.

2. What are the possible cases of sparse column?

In the design of any table there will be columns which can take NULL values and most of the rows in the column will be NULL due to nature of transaction. For ex: Spouse Name, If 50% are more are not married like me then you will have up to or more than 50% rows being null.

3.How to we efficiently handle the sparse column in terms of minimizing the space?

Use SPARSE keyword when you are defining a table to optimize the columns which contain high % of NULL Values.

In this blog, I have identified  two main areas where I could see the optimization of space.

1. Data space ( With and without SPARSE keyword )

2. Index space ( Filtered and Normal Index on top of sparse column defined using SPARSE keyword )

3. Combining the both ( 1 and 2 )

The Sparse Keyword optimization algorithm works differently for different data types.

Combining the Data + Index space usage

So, if you can combine the filtered index on top of a sparse column you can have a space saving of upto 70% * in certain cases.

But there are certain caveats like the SPARSE keyword cannot be used on certain data types and memory restrictions for updating the SPARSE columns.

For more information on the above points refer the MSDN blog :http://msdn.microsoft.com/en-us/library/cc280604.aspx

That’s all for now, please share your feedbacks and comments.

I Bing'ed  to get some facts about the Crop Circle Sightings around UK for the year 2009 and Also visited Met Office to collect few data points on how the weather was for the year 2009. All the weather was an Avg. Temperature from Min/Max of the day, Rainfall(mm) and also sunshine for that particular month in UK.

After Importing the data, I had to insert few images in my PowerPivot to make it interesting to present in visualization, So had to download and add those images in my PowerPivot.

The above image is posted as most of the developers in msdn ask Questions on how to upload image in PowerPivot.

Anyways, Once we had the data I used Powerview (Office 2013) to visualize the data.

Ok, Now we know that Aliens like April,June and July a lot ! and Barley and Wheat ! You know why barley and wheat :)

So, Whats so special in April,June and July Lets see that weather scorecard.

Conclusion :

1. Aliens like to relax in the sunshine

2. During sunshine they also like to prepare Juice ( Using Barley and Wheat ! :)) ! You guessed it right !

[Update] :: You can download the above PowerPivot from the link below

http://blogs.msdn.com/controlpanel/blogs/postlist.aspx?WeblogID=15553

Late night, Sitting under the A/C duct relaxing I could hear a grunted words from a near cubicle

About a strange requirement which requested by a customer. Keeping my curiosity at bay I thought its not my business …

It dint stop there ! I could here a disappointed words from another person the same bay .

Yet, I said to myself NO ! Let me not involve, but being a CONSULTING detective I couldn't leave with my curiosity and made a decision to look furtively and pretend

not being interested in the ongoing case.

I went to the last dark cubicle in the bay where I was hearing the sounds from and stood next to

Mr.SAM who was stuck to his laptop and dint seem to welcome the detective (you got it :) , who else but me).

He looked up to me and said " Hello Mr. MJ. The Sherlock homes of SQL Server "

Even tough it was sarcastic I stood there to see what's going on ?

And asked Mr. SAM about the other person who made a disappointed sound in the bay.

He said its  Mr.IG the new guy  from the Old Trafford.

Wasting no time, me and Mr.SAM went to Mr.IG bay to look what's the disappointed all about.

I said "Howdy ! What keeps you in bother ?" . All I could get was a grumpy face !

Hmmmm…..this sign is surely not welcoming !

Me :  So, tell me about the problem you are facing folks !

Mr. SAM:   Me and IG are working on a project which has requirement that the user should be able to see only the

data which he is eligible for , sort of an role based

Me:  hmmmm.. ( not jumping into any conclusion) , so what difficulty are you facing in that ?

Mr.IG:    We wanted to do this using @system_user function but the solution was rejected as they wanted to

Make it more secure !"

Mr.SAM :  Why is it so hard to find a solution for a simple problem ?

Me : OK ! Have you guys tested using symmetric keys to encrypt the data  to make it secure ?

Mr.SAM : " Hmmm…..But It will encrypt all the data in the column how can it be encrypted for a user ?"

Mr.IG :"Yes !

Me : Quoting the Sherlock homes " I am fairly familiar with all forms of secret writing, and am  myself the author of a trifling monograph upon the subject, in which I analyse one hundred  and sixty separate ciphers "

Mr.SAM :" Please, stop bragging and show us the solution"

Me : OK .

For readers : Symmetric keys are used to encrypt and as well decrypt the data.\

 Step 1 :

Creation of 2 logins which will simulate the 2 consultants ! One for samish and other for the new guy from old trafford IG.

Step 2 :

Creation of a database and table which will hold the encrypted data.

 Step 4 :

Creation of certificates one for each user which will be owned by the user which is done using the

Authorization keyword which are encrypted by the database master key.

Step 5:

Creation of symmetric keys which are encrypted by the certificates which were

created from the above statements using the Triple_DES encryption algorithm.

Step 6:

 Insert the data into the table encrypted the symmetric keys which are owned by the user.  Close all the symmetric keys after

Inserting the data.

Lets repeat the same for sam .

Step 6 : The Judgment step

As you can see , Mission Accomplished and case solved !

 [Update] : 3/8/2013

Below are the step by step guide

Step 1.Create scorecard has "sales territory" in the rows .

Step 2.Create the Report

The below is the report which I need to join with the above scorecard and the objective is to change the line graph when the user selects a "sales territory" from the above scorecard. For ex: when user selects "europe" change the report to show trend graph only for "europe" so on for "North America" .

• To achieve this need to have both the items (report and scorecard) from the same OLAP datasource.
• I have "Sales  Territory" in the background so that it can be linked with the scorecard.

3. Create a dashboard and create connection

Then you need to create an dashboard and add Report and the Scorecard to it.

Quick things to look out for :

1. The Report and Scorecard are connected using the Display Condition
2. You need to select the Row Member or Column Member
1. Row Member is the Attribute you have in your rows of the your scorecard.
2. Column Member is the Attribute you have in your columns of the scorecard.
3. You can also select the KPI itself if you need to connect it with the report.
3.  Select  the Member UniqueName from the above selected fields ( Row or  Column Member )

For our dashboard I am connecting "Sales Territory" which is RowMember (see the aboveScorecard)

So, Hover above the Row Member --> Member unique Name and drag and drop it into the Report which is an Analytic chart.

Once you are done with the above steps, You will observe the below pop-up which gives you the detailed information about the fields which are connected.

Observe :

1. Connect to : Destination field which needs to be connected
2. Source Value : The value gets passed from this field to the destination .

Next step, Deploy the above dashboard into SharePoint site. After that ...oopphh its all done.

See the below images, When clicked on "Europe" you will see  graph report for "Europe" and so on.

Thats all for now, share your feedbacks and comments below :)

Hierarchy is one of the mostly wide functionalities in any model. So, the usual question will be how do you RANK a City across ALL the cities for sales or Country across all the Countries for sales .

Lets say I need to RANK the City for the Country which it belongs and not across ALL the Cities!

How do you implement that in DAX ?

The image below will make it clear :

I have a hierarchy which goes from Country --->State --->City

RankAll Ranks all the City across ALL the cities . ( Shown in Green )

          Formula to achieve the same :

        =Calculate(RANKX(ALL(SalesFact),SalesFact[SumSales],,0,skip))

In the above example all the contexts are removed by using ALL function and hence Country  OR State context are discarded.

Now , we need to accomplish  these 2 requirements

1. 1.       Rank the City for which country it belongs to (Shown in Black)
2. 2.       Rank the city for which state it belongs to (Shown in Red)

To Achieve the requirement 1 :

RankCountry Ranks the city for the country which it belongs  to for ex : Germany has 4 cities and we need to rank ALL the cities in Germany and the same to United States .

              Formula to achieve the same :

              RankCountry:=Calculate(RANKX(ALLExcept(SalesFact,SalesFact[Country]),SalesFact[SumSales],,0,skip))

In the above example all the contexts are removed Except when you have Country in the any context which is achieved using  ALLExcept  function and hence when City is ranked the Country is context is taken into consideration.

To Achieve the requirement 2 :

 RankCountryState  Ranks the City for the state which it belongs ex : GermanyState1 has 3 cities and needs RANK these 3 cities.

                   Formula to achieve the same :

                   =Calculate(RANKX(ALLExcept(SalesFact,SalesFact[Country],SalesFact[State]),SalesFact[SumSales],,0,skip))

Post your comments why is it behaving like that with your explanation in the comments section :) 

• ALL(Table) will not apply for any of the contexts (Row,Query,Filter etc.)

•  ALLExcept (Table,Table[Column]) will apply the context only when Table[Column] is specified in any context (Row,Query,Filter etc)

•  ALL(Table[Column]) will not apply context filters for the specified columns i.e Table[Column] in any context (Row,Query,Filter etc)

Here is the structure of the "Doctors" Table : As you can observer the Spec and Hospital are from the same table from which i have created the measures which makes it intresting  !

The measures were sliced against Hospital , Spec column to observe the behaviour :

ALLSpec :=CALCULATE(countrows(),ALL(Doctors[Spec]))

ALLEXceptSpec :=CALCULATE(COUNTROWS(),ALLEXCEPT(Doctors,Doctors[Spec]))

ALLTable :=CALCULATE(countrows(),ALL(Doctors))

 Keep close attention to the the highlighted section

 As you can notice from the section :

fig  1. Hospital filter context doesnt change ALLEXceptSpec measure  as it will only change when the filter is on "Doctors[Spec]" Column .

fig  2. Spec filter context change ALLEXceptSpec measure  as said above and ALLSpec measure ignores the spec filter .

Here are few more links on the same :

http://technet.microsoft.com/en-us/library/ee634802.aspx

http://technet.microsoft.com/en-us/library/ee634795.aspx