There are a number of useful preferences you might want to customise.

From the Menu you can access Preferences. These are loaded every time you open corpkit. Most defaults are sensible, but a part of the philosophy of the tool is that any arbitrary value should be customisable.


Name Function
Automatically check updates When selected, corpkit goes online on startup, looks for newer versions, and requests to install them if available.
Truncate concordance lines By default, corpkit will ask you to truncate concordance results if there are more than 1000. You can change that number here, or just hit No at the prompt to load them all.
Truncate spreadsheets Unlike concordances, spreadsheets are automatically truncated at 9999 unique results. This value can be altered here.
CoreNLP memory allocation The CoreNLP parser requires a lot of memory, especially if you have very long sentences or if you are using referent tracking in long documents. While the default value should be sufficient for parsing well-edited text, non-standard material might need an increase. Your computer only has finite amounts of memory, however, so performance problems may result if you raise this number.
Spreadsheet cell width If you have a lot of subcorpora, you may want to reduce the width of spreadsheet cells to see more results.
Spreadsheet row header width If you have long corpus names, you may want to increase this dimension to allow them to be seen
P value When editing interrogations, some options (sorting by increase/decrease, or selecting Keep stats) will cause p values to be generated, representing the likelihood that variations between subcorpora are caused by chance. If you change this number and use the Remove above p option, anything result higher than the threshold will be automatically removed
CoreNLP path corpkit will download CoreNLP to your home directory. If you move it, you can tell corpkit where you moved it to.