Tableau is a drag-and-drop pivot chart and graph builder. It connects to a live datasource and performs a series of extract statements to build graphs. Tableau builds a cache history to reduce load on the datasource. No ETL functions are available.
In practice, Tableau needs to be connected to a screaming fast data warehouse. Large datasets may still need summarization for adequate performance. Using clean data with descriptive column names will save the need to hide or rename columns in Tableau.
Tableau determines whether the data can be shown as a graph or chart and opts to display a chart. The use of trellis charts is very helpful across multiple dimensions. It is very intelligent in chart selection.
See for yourself in the product tour
Con:
- Updates require a series of re-queries against the datasource.
- No ETL functionality.
- Large data sets are clearly a problem. My demo of 128,000 rows slowed down as I added more than 4 dimensions.
Pro:
- Automatically switches between chart and graph, opting for a graph where possible.
- Trellis charts look great and are very effective.
- Adding color, choosing sort options, and tweaking the graph is generally easy and fun.
Related posts: