This article identifies some key general strategies for working efficiently in Displayr:
- Avoid writing code
- Duplicate, then modify
- Understand variable set structures
- Use wildcards in find / replace
- Tidy as you go
- Comment as you go
- Minimize the size of the dependency graph
- Update your data
- Do your data prep in Displayr
Avoid writing code
Duplicate, then modify
The basic workflow in Displayr is that you create something that's great and then duplicate it, and further modify it. For more information, see Automatic Updating, Duplicate & Modify, and Extensibility.
Understand variable set structures
Variables are always grouped into sets of 1 or more variables. The resulting variable sets have a structure that dictates how they are analyzed (see Variable Sets).
Use wildcards in search / replace
Consider the variable sets below. Let's say you wanted to delete everything up until and including the space after the dash. This can be done using wildcards.
The process is:
- In the Search box, type "*- " (escluding the quotation marks) and press This Document.
- Press Advanced and check Use wildcards. This means that anything prior to the dash will be identified.
- Press the Replace tab, leaving Replace with blank.
- Press Replace all.
Alternatively, if you wanted to add something (e.g., the word POST) to the end of every label, you would use:
Tidy as you go
Make your labels and titles neat and correct, so you can export without having to re-do them. Also, keep in mind that as you will regularly be duplicating, it's more efficient to tidy once and then duplicate, rather than the other way around.
Comment as you go
Displayr allows you to add commentary to your page while doing your analysis. This is much more efficient than first doing analysis, and then trying to remember your insights again.
Minimize the size of the dependency graph
If you aren't a bit strategic about what you link and what you don't link, a Displayr document can grind to a halt because the dependency graph gets so big that a lot of time is spent checking things whenever you change anything
For example, it can be a good idea to do data preparation in one document, to export it to the Displayr cloud drive, and then import the tidied document into a dashboard.
Update your data
Once you have created a document, you can update it with new data. Your tables, visualizations, and other calculations will automatically update (unless you specify otherwise). To update, click on the data set name in the Data sets tree and then click the Update button in the object inspector.
Do your data prep in Displayr
For basic data cleaning and preparation work, it's much, much, faster to do this in Displayr than, say, R or SPSS. And, it is superior from a quality and reproducibility perspective as well.
It can be sensible to do it in a different document.