KNIME

KNIME has been a great tool for my data analysis and transformation needs. I can easily grab and combine data from different sources, without having to code or coding just a few lines (i.e. some rules, small snippets to string manipulation or to input math formulas). Performance wise, things look great: I can specify the maximum memory usage inside “knime.ini” using a text editor (the default is just 1 GB) and big workflows run in a couple minutes.

For urban and regional planning, I think KNIME is an ally, as it allows me to fetch and process census data in a manner that is easily to demonstrate, document and replicate: no more fiddling with Excel spreadsheets; no more manually redoing tasks and losing time because some variables changed; no more sluggish VBA code to write.

I would love to see xBase support out of box, because ESRI Shapefiles are predominant for those that, as myself, work with cartographic representation and GIS. The looping capabilities of KNIME would make very easy to write metadata and dictionary files for big sets of files.

The community is wonderful too: many examples readily downloadable and good forums. I had all my doubts solved very quickly reading the discussions, exploring previously built workflows and reading the docs. There are official videos introducing key features of KNIME.

The KNIME ecosystem is affordable: the server part is paid, but the modeler is not. One can built complex workflows without trouble. The server looks specially useful for big companies that need collaboration and cloud based solutions. Even without the server part, if the workflow is built accordingly, being tied to the workspace, one can export an archive and share it with colleagues and coworkers.

KNIME files also use open standards, like XML and Zip, making backup and snapshot a breeze. I use Back In Time to automatically record a snapshot every 30 minutes and can audit the changes for every group, workflow and node.