CSMDM21 - Data Analytics and Mining Practical 1: KNIME Basic I/O
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
MSc Data Science and Advanced Computing
CSMDM21 - Data Analytics and Mining
Practical 1: KNIME Basic I/O
A: Data Input
1. Create a KNIME workflow and use the node File Reader to read the data file directly from the URL: http://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.data
B: Adding Column Headers
The .data file does not contain column headers, which can be found in the corresponding files with extension .names: http://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.names
It is a good practice to add column headers to make data exploration more user friendly. This can be achieved in different ways.
1. The simplest method is to change the column headers directly in the node File Reader.
2. Another simple way is to use the node Column Rename, which is specifically meant for this purpose. However, in both cases the process is tedious as the data contains many (15) attributes.
3. An advanced method is to build and use a dictionary to change the column headers of a table to the desired values. Create a text file with the list of desired column headers (one header per line) or download adult_headers.txt from Blackboard. Read the column header file in the workflow and use it to create a dictionary old_name vs new_name for the column headers. For thistask you can use the nodes: Extract Column Header, Transpose, RowID, Joiner, and Insert Column Header.
C: Result Output
1. Use the node Interactive Table (local) to view the result table.
2. Use the node CSV Writer to write the result table to a new file.
Solutions:
Sample solutions to these exercises are available on Blackboard (Bb) in two forms: images of the workflows and the actual KNIME workflows (as a single zip archive). You should first try to build your own workflows for each exercise. During the practical session, you may use the images to see the proposed solutions and to reconstruct them. During or at the end of the practical session, the archives with the actual KNIME workflows will be made available for you to import and test them. These are KNIME archives (file extension ". knwf ") that can be imported in KNIME. Note that after importing the KNIME archives, you may need to change the file locations of the source data file and the output file destination folder.
2023-09-26