Function update

[Tess-LaCoil]

Update to a collection of functions:
assign_data

• Now takes either dedicated named vectors or a dataset with appropriately named columns
• Throws errors when the data does not match the required form
• Throws an error when the requested model does not exist
• Checks sizes for vectors in the created data list
• Checks for existence of elements of created data list

run

• Now checks if the chosen model exists in the package

Selection of new functions:
extract_estimates

• New function that takes in a model name, matching fit, and returns a list of data frames corresponding to measurement, individual, population (if multi-individual), and error estimated parameters.
• Error handling to address Issue Error handling and testing for extact_estimates function #34
• Error structure: provided fit has to exist and must match the requested model parameter names, and the model must exist.

plot_de_pieces and ggplot_de_pieces

• First is a wrapper for the second that does some data wrangling like find the start and end values of $Y$ for each individual.
• New function that produces a figure of $Y$ against $dY/dt$ based on the chosen model and individual parameters.
• Fits a piece to each individual from $Y_i(0)$ to $Y_i(final)$, parameterised by $\theta_i$
• Has some limited customisability.

model_des

• Function that returns the differential equation as a function for each of the four included models.
• Checks if the model is available in the package.

model_names

• Function that returns a vector containing the names of available models.

model_pars

• Function that returns a list of character vectors for each available model
• Character vectors give parameter names for each level in the model
• Checks if the model is available in the package.

Test coverage
New test scripts for assign_data, extract_estimates, and plot_de_pieces.

[Tess-LaCoil]

Tests failed because of the load order of dplyr and stats producing warnings, see lines 70, 71 in the macos checks for an example.

[Tess-LaCoil]

I am testing whether removing stats from the imports and using importFrom for the quantile function in the extract_estimates code works to fix this.

[dfalster]

Great work here Tess!

I agree with removing stats from imports, much better to just call in specific functions than entire package.