Displayr

Currently, if a user's API key updates in a data collection product that has an API connection with Displayr, then there's no way to update this within Displayr proper. It has to be done by editing the underlying .q file by Support.

Broader customer issue to solve: When updating a data file i.e. for a tracking study, sometimes the new data file might contain additional variables (in case of a multi-response question) or values (in case of a single-response question). When the data is imported, it provides the user with a warning that there are changes in the data file and then when the data is updated it places additional variables outside of a variable set, causing the user to merge the variable manually with the set. It would be helpful to users if they had more control over this, i.e. we can bring up a screen that says "this is a new variable in the data file, we think it belongs to this question, please confirm if you would like to group it with this question or if you would like to keep it as a separate variable"

The cloud drive is used to store data and model outputs generated from that data. A user may have many files and it is difficult to recall the names of all the files and cumbersome to type/copy and paste them from a separate tab that shows the files in the cloud drive. Use case is a user has a dashboard to view and/or analyse specific data/outputs from their cloud drive. They can see what data is available on the hard cloud drive and load the one they want without having to open the cloud drive and do a manual search in a separate tab. E.g. a combo box control can be populated with the filenames that exist on the cloud drive. It can update when a new file appears. Right now there needs to be a static R output that requires manually typing the file name to add to the control. Suggested implementation flipAPI:::QListFiles(pattern = "*")

Data quality is, at the moment, spread across a number of different features in Displayr which have to be run individually depending on what type of check you want to run. It would be good to have a central location for data quality checks, including all the "standard" scripts and features, and automate data quality as much as possible. Consider a single feature with an object inspector that allows the user to specify which variable(s) to use for speeding checks, which questions should be included in flatlining or patterning checks, what tests to run over open-ended data to identify potential bot responses, identification of duplicate cases, or inconsistent data, all with user-specified tolerances. Produce a filter variable for each test, and an additional filter variable that identifies cases that exceed a user-specified amount of tolerable "errors." Finally, a function to automatically delete from the file any case that fits the "final" filter and updates accordingly if there's added data in the source file (or if the settings change).

Users often face challenges when importing updated data files that do not exactly match the original data file. The Data Difference warning is useful only in as much as it says where there are problems, but it does not provide any help in fixing those problems. In both Q and Displayr, we only indicate in the UI that there IS a problem, and in both cases offer only limited assistance in helping users identify and ameliorate changes. The Data Difference warning dialog should be replaced with a dialog similar to the merging dialog in Q, with options to: State that a new variable should be part of an existing variable set, to prevent it "sitting around" on its own in the Data Sources and for the user to discover it. Allow the user to accept or reject differences between variables, and whether to merge them or create a new variable. Many other potential situations exist - the above are a couple of quick examples.

Snowflake is rolling out mandatory MFA, and the user would like to modify their connection string so that they do not have to manually approve each connection request.

For safety and security reasons, we need an option to upload data directly from a SharePoint folder. This feature would allow us to securely manage our data without needing to make links publicly available.

Broader Customer Issue to Solve: Data comes in a number of styles and formats, currently Displayr runs through a series of checks to understand the best Variable Structure to assign to variables/questions. With poorly structured data it could get this wrong and then default to not structuring variables into questions (this is particularly true for Binary Multi's, Binary Grids, Nominal Multi's). This makes it difficult for users as they do not always know how to combine the split out variables and adds A LOT of processing time to their reporting. Potential Solution We can make it easier for users by bringing up a screen upon data import that allows users to select the most appropriate question type when Displayr is "unsure" on what to do with it.

The interface for Stacking data and Combining By Case in Displayr is not as user-friendly as in Q. In Q, users can drag and drop single or multiple variables at once making it very simple to align variables. In Displayr, this process can be much more difficult when there are changes (e.g., Variable Name, Variable Label, Response List, etc.) to the survey and when there are more than two data sets. An easy-to-use interface for Stacking data and Combining By Case in Displayr would be a major quality-of-life improvement. Displayr + > Data > Data Set > Combine > By Case + > Data > Data Set > Stack Q Tools > Merge Data Files > Add New Case Tools > Stack SPSS .sav File...

Add the functionality to strip HTML from labels when importing data sets into Displayr, matching the functionality of Q.