Visualization of big date/time-related data

[krizz86]

Hello everybody,

I hope this is the right place to ask for support for my Power BI problem.

Desired Visualization 

The system measures every 15 minutes the power consumption so that I got a lot of data points over the year and years. Data are saved in a Excel file with following columns: Data ID, data point date/time stamp and consumption. All data points should be evaluated (per year, month, day).

So I started to set up my Power BI report and I got a good visualization if the column data ID is on x-axis and consumption on y-axis. I can filter the data according to year, month and day and the visualization adapts accordingly. However, the information of the x-axis = data ID is not helpful since this does not include any time information.

Therefore, I have started to create calender and time tables and to link them with my data table. In priciple, my approach seems to work but the visualization is too big to show all data points due to the representation of each individual x-values.

My question:

Is there any possiblity to combine both visualizations? I want to visualize all data points which I filter from the data table, e.g a year or a month, but with date/time information on the x-axis instead of the data ID.

I hope I could explain my Power BI problem acceptable for you und that you can help me to solve my problem.

Best regards,

Christian

[Anonymous · Accepted solution]

Hi @krizz86 ,

Thanks danextian  for the quick reply. I have some other thoughts to add:

(1) You can create a column.

date-time = [Data Point Date]&" " &[Time Stamp]

Change data type.

(2) Create a line chart. Use the [date-time] column as the x-axis and change the x-axis type to Continuous.  Such an x-axis type avoids scrollbars on the x-axis of the visual.

As for the Data ID, you can display it in the line chart as a tooltip or as a second y-axis or legend.

Best Regards,

Neeko Tang

If this post  helps, then please consider Accept it as the solution  to help the other members find it more quickly. 

[danextian]

Hi @krizz86 

If you have a column that contains both the date and timestamp, then this is possible but this will of course increase the cardinality of  your data and depending on the size may lead to a degraded performance. However, using a hierarchy and expanding it disables a continous axis type.  Alternatively, you can try to find a balance between showing lesser datapoints and being able to use time by rounding your time to the nearest hour or hours -for example 12:00, 12:15 and 12:30 and 12:45 are  all grouped into 12:00.

[krizz86]

Hello @danextian, thanks for your support.

You are right that my table have a column containing date and time stamp.

I have tried your proposed way to use hierarchy for a non but nevertheless I have come to the same result as before. So I have my data ID on x-axis and visualization looks good or I have date and time as information on x-axis but I can only see a small number of data points. Maybe my attemp was not your proposed way. Maybe you have a link to an introduction which I could use?