Data transformation is a term that is used to describe the conversion of data from a source into a destination. Destination data applies to meta data (data about the data) and the data transformation process takes place in two basic steps. It sounds simple in theory, but data transformation is often quite intricate and may require one to many and many to one transformation rules.
Data Transformation Steps
Data transformation generally takes place in two steps. The first step involves mapping of data and data elements from the source to the destination along with any data transformation that may occur. The second step is the code generation that will create the actual data transformation program or map. The code generation aspect of data transformation will actually creates an usable program that can be installed on a computer system. In addition, during the code generation portion, computer languages that are easy to maintain can be created.
Data Transformation and XML
Data transformation can be done in a few different languages. There are a handful of data transformation languages, all of them having different uses and requirements for grammar. The grammar is not unlike Backus-Naur Form or BNF. Each of the data transformation languages varies in its purpose, cost, transformation level of value. Two of the more popular data transformation languages are XSLT, which is an XML data transformation language and TXL, which is a prototype language that is used in and for data transformation.
Data Transformation and C++
Data transformation is actually quite difficult and many people struggle immensely with it. One of the biggest problems is with C++. In this form of data transformation the problem usually lies with the unstructured preprocessor directives. These are preprocessor directives that do not have blocks of code with simple grammar descriptions, making the transformation quite hard. When there are problems such as this, the DMS Solutions Reengineering Toolkit is usually quite helpful.
Data Transformation and Distribution
Data transformation is not something that is for everyone and it is very complex based on the data that you are trying to transform as well as the language that is being used in the process. The idea behind the data transformation is to be sure that it has a normal distribution, and this required the need to understand transformation to linearity, kurtosis, and skewness, all which contribute to the normal distribution. There are many different techniques that are used to make sure that there is normal distribution such as logarithm, square root, reciprocal, and cube root. Data transformation is simply a difficult topic that many people are never quite able to master they way that they would like because the grammar is hard to get just right, as is the normal distribution.