The term transformation is used in high-frequency and information technology in the context of converting signals from a spatial plane to a frequency plane and in connection withdata structures and data transformation.
The transformation from a spatial domain to a frequency domain
A Spatial Domain is a spatial domain. In terms of an image space, it is a two- or three-dimensional representation in which each pixel is uniquely determined by its position in the plane or in space. Changes to the images are applied directly to the pixel values. The number of pixels determines the image resolution.
In a transformation, the pixel values can be transformed from a time-based representation to a frequency domain using mathematical operators. The pixel values of the image then appear frequency related.
Various methods are available for these transformations from a time domain to a frequency domain, such as the Fourier transform or the Laplace transform.
The transformation of data structures
The data transformation is the central part in an Extract, Transform, Load( ETL). Here, the data from the source system is transformed into the form required by the target system. When transferring data from one system to another, i.e. in a data flow process, it is almost always necessary to perform a transformation.
- Data type conversion: the data is converted from the data type of the source system to the data type of the target system. For example, a 1 can also be converted to a "Yes" and a 0 to a "No".
- Simple transformations: For example, only the first 10 characters can be passed from a text field.
- Simple formulas: E.g. the profit margin is calculated, which is not directly available in the source system, but is calculated from the formula "sales price - contribution margin".
- Mapping: Data that is not related in the source systems is assigned to each other via mapping. For example, the geographical position of a customer's place of residence is assigned to the customer.
- Integration: Data from different source systems are reconciled and linked, e.g. data from the accounting system and the order processing system.
- Quality assurance: Data ischecked for quality and corrected. For example, addresses entered manually in the source system are compared with data from official address directories.
- Filtering: Unnecessary data is filtered out. E.g. completed orders are filtered out of the order backlog data.
- Aggregation: Data is summarized to reduce the amount of data and to bring it to the required (lower) granularity. E.g.: detail items of a booking are aggregated to the main items.
- Distribution: Data is distributed to detail items. E.g. rough plan figures are distributed to individual months and products.
- Relative and statistical calculations: Based on the data from the source system, comparative values are calculated, e.g. the percentage of satisfied customers.
- KPI(Key Performance Indicator) calculations: KPI values with target and actual comparisons and states are calculated from the data of the source system.
- Structural change: Data is often required in the target system in a different structure than in the source system. For example, relational data is converted into an OLAP structure.