Proposed Overhaul of the Package and Functionality

[chantelwetzel-noaa]

Overview

Now that the 2021 assessment cycle is coming to a close, I would like to propose an overhaul of the package to create additional functionality and improve the ease of use. The existing code and package is similar to the code created 10+ years ago. Updating the code to improve usability and readability of the code if long overdue.

Intended Outcome

The goal would be to create robust function calls that allow the user to pull, process, and preview data with just a few lines of code.

Proposed Actions

• Clean-up existing branches  #50
• Add functionality in the data pulling to access non-standard survey data
• • PullBio.fn enhancement / documentation of filtering #45
• • Lifestage samples and their origin #43
• Pull data from multiple surveys with just one function call #17 (Pull data from multiple surveys in one call)
• Create wrapper functions to bundle associated function calls
• • Create wrapper to pull both catch and biological data together #51
• • Create wrapper function to calculate design-based indices and figures #52
• • Create wrapper to produce length and age compositions with all figures #53
• Add new data visualizations
• • add function to show presence/absence by depth bin? #31
• • Add figures showing the distribution of lengths from aged fish #54
• • Add length/age frequency figures similar to those created by r4ss #55
• • Move towards using more {ggplot2} graphics #34
• Apply consistent code syntax and style #56

Future Ideas

An area to table any of the proposed actions if needed.

[kellijohnson-NOAA]

@chantelwetzel-noaa I love the organization and clear statement of goals here. I am wondering if we can edit the "Intended Outcome" to be

create robust function calls that allow the user to pull, process, and preview data with just a few lines of code

I think this is a great mission statement and we could organize functions into these three groups

• pull
• process
• preview

I am even wondering if we should use the first four letters as the beginning of function names, e.g., pull_base pull_catch. You might not want that level of prescription in the naming structure though. And instead, we could use the @family capabilities of roxygen to categorize and group functions together.

[chantelwetzel-noaa]

@kellijohnson-NOAA This issue was written with you in mind! My thinking so far was to create clear grouping of functions via the function name (e.g. pull_catch()). I have not fully worked out the 'right' name grouping for the different processes but the ones you suggested are in line with my thoughts. Here is so rough ideas of what I was thinking:

• pull_: download various data products from data warehouse
• process_: calculate data products (e.g., indices, composition data)
• plot_, preview_: create data plots
• get_, 'do_': wrapper function that will bundle steps (e.g., process_ and plot for a data type/s, pull_ for both catch and biological samples)
• other functions: need to identify a naming structure for worker functions that are called within functions but are not typically called independently.

I know renaming functions can create challenges for established users but hopefully with good documentation and clear naming structure and flow will alleviate these challenges.