Power BI Practice: Concatenation generally does not improve dataset performance

Scenario:

Given meta data:

column_name, data_example, valid_profile

IoT GUID, 48196321-3809-EC11-883D-0022489A2..., 100%

IoT DateTime, 21/05/2022 18:59:25, 100%

IoT ID, 10000010000, 100%

The IOT ID columns are unique to each row in query. You need to analyze I0T events by the hour and day of the year. The solution must improve dataset performance.

Solution: You create a custom column that concatenates the I0T GUID column and the IoT ID column and then delete the IoT GUID and IoT ID columns. Does this meet the goal?

Answer: No, this solution does not meet the goal of improving dataset performance for analyzing IoT events by the hour and day of the year. Concatenating the IoT GUID and IoT ID columns into a custom column and then deleting the original columns does not provide any performance benefit for time-based analysis.

Instead, you should focus on optimizing the datetime column for your analysis.

Step-by-step guide

[1] Extract Date and Time Components:

Create separate columns for the date and time parts from the IoT DateTime column. This will make it easier to analyze events by hour and day.

• Date Column: Extract the date part.

• Hour Column: Extract the hour part.

[2] Indexing: Ensure that the datetime column is indexed if you're using a database that supports indexing. This can significantly speed up queries.

[3] Aggregation: Pre-aggregate the data by hour and day if possible. This reduces the amount of data that needs to be processed during analysis.

[4] Remove Unnecessary Columns: If the IoT GUID and IoT ID columns are not needed for your analysis, you can remove them to reduce dataset size.

By focusing on these steps, you can improve the performance of your dataset for time-based analysis.