Aggregation is the method of collecting data based on a condition and performing analytics on the data. Aggregation is very important to make sense of data of all sizes, as just having raw records of data is not that useful for most use cases.
For example, if you look at the following table and then the aggregated view, it is obvious that just raw records do not help you understand the data.
Imagine a table containing one temperature measurement per day for every city in the world for five years.
Shown in the following is a table containing records of average temperature per day per city:
|
City |
Date | Temperature |
| Boston | 12/23/2016 | 32 |
| New York | 12/24/2016 | 36 |
| Boston | 12/24/2016 | 30 |
| Philadelphia | 12/25/2016 | 34 |
| Boston | 12/25/2016 | 28 |
If we want to compute the average temperature per city for all the days we have measurements for in the above table, we can see results which look similar to the following table:
| City | Average Temperature |
| Boston | 30 - (32 + 30 + 28)/3 |
| New York | 36 |
| Philadelphia | 34 |