Let's say you have uploaded data in a file with 3 columns: one for IDs, one for the main text data, and the last column for a rating. When you create your project you choose which column contains the main text data. But the other columns are also part of the project and they are available for filtering. Every time you execute explore by pressing the explore button you may optionally also specify filters.
The filter section is located above the main display of your exploration results. In this section you can apply filters from the other columns of your data file. Columns available for filtering are those that contain data that is similar across rows. So in our example, the ID column will not be available for filtering since every row will have a unique ID, and the main text row is used for analysis so that won't be available either, but the rating column will be available since it has only 5 distinct values - 1 through 5 - for all rows.
To apply a filter you choose the filter column from the drop down menu and then click the values you would like to filter on; in the screen shot below I have chosen the ratings one and two. You can set up filters using several columns at the same time.
After clicking 'Add filter' you will see this:
When your filter setup is complete you can re-explore your project by clicking the Explore button. At this point only the data that passes the filter configuration will take part in the analysis. Pinned topics and groups will be present in the results even if they no longer have any associated texts. If your filter configuration is too constrained so that there are no matching texts, the system will tell you.