VA Session 8 & 9 Alteryx ^ Flashcards
Data Transformation
- adding or modifying variables
- e.g. computing variable C from A & B or applying log-transformation on single variable
Data Manipulation
- modification of information to make it easier to read or more structured
- Data transformation (adding or modifying values) & filtering, sorting, aggregating
- e.g. data values sorted in alphabetical order, making it easier to find individual entries
Data Modelling
Applying statistical models or
Way of identifying & mapping data used in business processes & relationship (data engineers)
Dataset
collection of values (numbers or strings)
Values
belongs to a variable or observation
Variable
contains all values that measure same underlying attribute (e.g. height) across units -> columns
Observation
contains all values measured on same unit (e.g. person, day) -> rows
Alteryx Definition
- low-code no-code data blending & transformation tool (making data accessible to anyone)
- Goal: clean data where variables being columns & observations rows (Alteryx & Tableau prefer this data structure)
Workflow design
Should be designed in structured & well-documented manner (you later & colleagues should understand)
Workflow Design - Methods
o Colors
o Containers
o Use of space (unlimited)
o Input on left hand side, outputs to right
o Documentations incl. comments
o Workflow tests (message = I and tests = + / - sign)
Difference input data & input text
- Input Data: reads data into workflow (select a file or connect to a database)
- Input Text: manually enter text to create small data files for input (useful in testing while building workflow) -> data saved within the workflow -> share the workflow without having to provide an input data filter
Browse Tool
- to view data from a connected tool & data Profile information
- for multiple columns in single holistic view, or for 1 column
- information on data type, number of records, data quality, and a variety of statistics.
Cross Tab vs Transpose
- Cross Tab: from long into wide format
- Transpose: from wide into long format