When we have a series of quantitative data, such as latency in milliseconds, for the same operation (measured over a number of executions), we can observe a number of things. First, and the most obvious, are the minimum and maximum values in the data. Let us take an example data set to analyze further:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
We can see that the minimum latency here is 19 ms, whereas the maximum latency is 172 ms. We can also observe that the average latency here is about 40 ms. Let us sort this data in ascending order:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The center element of the preceding data set, that is, the ninth element (value 23), is considered the median of the data set. It is noteworthy that the median is a better representative of the center of the data than the average or mean. The center element of the left-half of the...