Is there a term for data based on calculations of raw data?

Question

I'm trying to find out if there is a specific term for describing data created from computations based on raw data.

For example, I have the two data tables below. I'd like to know the term used to describe table #2. Thank you!

Data Table #1: Raw Data Based On Sales Transactions

|**TimeStamp**|-|**TransactionTotal**|
--------------------------------------
|   09:30     |-|       100.00       |
--------------------------------------
|   09:31     |-|       50.00        |
--------------------------------------
|   09:35     |-|       10.00        |
--------------------------------------
|   10:12     |-|       200.00       |
--------------------------------------
|   10:18     |-|       100.00       |

Data Table #2: Calculations based on Sales Transactions

------------------------------------------------------
|   Avg Transaction Value           |-|       92.00  |
------------------------------------------------------
|   Transaction Per Hour (09am)     |-|       3.0    |
------------------------------------------------------
|   Transaction Per Hour (10am)     |-|       2.0    |
------------------------------------------------------
|   Total Transactions              |-|       5.0    |
------------------------------------------------------
|   Avg Transaction Per Hour        |-|       2.5    |

Welcome to DataScienceSE. This is usually called feature engineering afaik, — Erwan, Aug 24 '22 at 18:25

score 2 · Accepted Answer · answered Aug 24 '22 at 21:13

In your specific case there, I would call them descriptive statistics of the raw data.

In the context of feeding those statistics into a model (treating them as features), you could call them manually generated features $\rightarrow$ the result of feature engineering.

Is there a term for data based on calculations of raw data?

1 Answers1