This is the RefinePro knowledge base about OpenRefine. We build it over the years, and keep adding to it. From great tutorials and how-to, to handy GREL expressions and links to external resources, you will find here one of the most comprehensive list of resources to learn OpenRefine.

For a comprehensive documentation you should refer to the official OpenRefine wiki.

Don't where to get started? Search for a specific function below, or read our most popular article from the right side menu.

28.7.11

remove " (quotation) mark

Hard time removing the " (quote sign) from your expression. Instead of quoting your quote mark with double quote, do it with simple like this :

22.7.11

Remove or replace a specific character in a column

You want to remove a space or a specific character from your column like the sign # before some number.

21.7.11

merge 2 columns that have both blank cells

The basic function to merge or concatenate data from two different column is the following:
cells["col1"].value + " " + cells["col2"].value




19.7.11

Add a space to postal code (splitByLength and Merge function)

This short tips explains how to convert postal code store on 6 characters to 7 by adding a space after 3 digits. We will use splitByLength (see related video) and merge multiple column into one functions.

18.7.11

Compare values from two columns

To compare strings from 2 differents column and present the results in a third one, use the following expression:

5.7.11

Move data from a column to an other

Some data are messy and column can be a mix of different type of data. For example if you work on an address database, you might have a ZIP code column and a State column. And some ZIP code can be stored in the State Column.

2.7.11

Faceting with Freebase Gridworks

Freebase is the name of google refine before google took control of the solution. The two following videos present how to facet (filter) in google refine. The interface and options did not evolve too much over the time and version making those videos still up to date.