Implement GBA code suggestions

[wendtke]

@iqis I met with @geanders today. She recommended making the following changes:

• 1. restructure expected study directory to separate folders (e.g., subject_1/task_1 and task_2 and task_....) rather than file name with subject_task
• 2. Use importFrom magrittr %>% in roxygen notes to specify pipe
• 3. Don’t put suppressWarnings() within function; instead, use purrr::quietly or purrr::safely
• 4. For exported functions, put an example in roxygen notes (deferred, pending better sample data)
• 5. Define precise parameters (e.g., a character string that gives path to... rather than path only)
• 6. user-friendly output of print(psyphr_workbook); see Generics for psyphr_workbook #43

[wendtke]

I started the second TODO.

[iqis]

on suppressWarnings(): I admit it was a bad habit.

[iqis]

Do we have a reason on 1? I feel a flat structure is much easier for ready_study() to parse file name sections as ids.

[wendtke]

Do we have a reason on 1? I feel a flat structure is much easier for ready_study() to parse file name sections as ids.

She recommended that not all metadata (subject id and task) exist in file name. Plus, most labs follow the "study/subject_N/task_x/y/z" file directory structure.

[iqis]

Do we have a reason on 1? I feel a flat structure is much easier for ready_study() to parse file name sections as ids.

She recommended that not all metadata (subject id and task) exist in file name. Plus, most labs follow the "study/subject_N/task_x/y/z" file directory structure.

subject and task ids will be prepended to the file name if they exist in the form of a directory structure. For example,subject1/task2/subtask1.xlsx will be turned into subject1_task2_subtask1.xlsx. The only problem is that there cannot be _'s in directory or file names.

The reverse conversion can also be done, and I think psyphr can accommodate both.

Done with commit 6bea378
1 done with commit 75a097f

[iqis]

4 deferred pending more stable API and better sample data.