Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
9 years ago

Joining dates at different granularities

Hi

 

I have two SQL queries from different databases.

 

Database A contains hourly price data

Database B contains daily price data

 

I need to create a chart showing the total hourly price by multiplying the hourly price with the daily price. So the daily price effectively needs to be repeated 24 times a day so that it can be multiplied to the hourly price.

 

I have a calendartable but I can't figure out how to join them but to be able to use hourly granularity in tables and charts.

 

thanks very much in advance

  

 DATABASE A - hourly price

 

DATETIME                        VALUE

01/01/2017 00:00:00           45.45

01/01/2017 01:00:00           34.56

01/01/2017 02:00:00           43.34

 

DATABASE B - daily price

 

DATETIME                          VALUE

01/01/2017   00:00:00         15.67

02/01/2017   00:00:00         13.56

 

Need Help

2 Replies

  • Phil_Seamark's avatar
    Phil_Seamark
    Microsoft Employee
    9 years ago

    Hi Anonymous

     

    I think I follow what you need.

     

    What I did was add a new column to the HourlyPrice table that only contained the date.  I did this using the Add Date column in the query editor

     

     

    This allowed me to create a relationship between the HourlyPrice and DailyPrice tables based on my new HourlyDate.Date column and the date column already in the Daily table

     

     

    With this in place I added this calculated measure to the Daily table

     

    Daily Price = CALCULATE(SUM(Daily[Value]))

    Which I then used in a new calculated column on the Hourly table

     

    Hour x Daily = CALCULATE('Daily'[Daily Price]) *'HourlyPrice'[Value]

    This is the result based on your sample data.

     

     

    Are we getting close?

     

     

     

    • Anonymous's avatar
      Anonymous
      Not applicable
      9 years ago

      Hi Phil

       

      Thanks for the quick reply. I think it's almost there but it's not quite getting to where I wanted. Perhaps my example was too simple so wasn't quite covering what was needed

       

      I've uploaded some data here over three tabs

       

      https://www.dropbox.com/s/wusqsxeyt3tyeh2/powerBImixeddates.xlsx?dl=0

       

      What I need to end up with an hourly data series  to create a spark spread ( the theoretical profit of running a power plant in case you are interested!)

       

      The formula is: HourlyPowerPrice - ((DailyGas[VALUE] + DailyCarbon[VALUE]*0.202)/0.55

       

      I tried to apply your suggestions but am hitting a problem. I managed to relate the charts via the date as suggested.

       

      The daily values are different for each day but what I am getting is then an average of the daily values over the whole time series rather than a distinct price for each day.

       

      Does that make sense?

       

      I am very grateful for any pointers