In npjc/growr: High level functions to help summarise growth curve fits.

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

growr

growr helps you with the pieces of growth curve analysis.

1. Preprocess growth curve data: preprocess()
2. Summarise data with a family of functions: fit_*()
   • use built-in or your own metric_*() functions:
   • use built-in or your own models via fit_nls(formula = ) 3

High Level Interface

Currently, growr has high level interface that is meant to be simple with reasonable defaults:

1. preprocess() transform raw measures data into values suitable for summarization.
2. summarise_fit() summarise a fit method into growth-curve metrics such as the fit's $A$, $\mu$, $\lambda$.
3. add_grouping() and add_groupings() provide a convenient way to add relationships between sets of wells of a microtitre plate. Reference and target well designations can then be used to rescale summary metrics within groups (with for example, the provided relatively()).

lower level functions

growr also exports an extensive set of models that make up a low level interface. Nearly all these functions have previously been used in growth curve analyses. The majority of this functionality is not new. It reflects a collection from various public source from this author's exploration. Citations provided in documentation of the individual methods.

Example

Preprocessing

preprocess() takes a vector of values and returns a transformed version. The details of the transformation depend on the argument choices but follows the preprocessing steps laid out in the precog paper.

library(tidyverse)
library(growr)

data <- mtpview::mtp_example3
data
data %>% 
  group_by(well) %>% 
  mutate(measure_pp = preprocess(measure))

Summarising fit

# well_summaries <- data %>% 
#   group_by(well) %>% 
#   mutate(measure_pp = preprocess(measure)) %>% 
#   summarise(metrics = list(summarise_fit(x = runtime, y = measure_pp))) %>% 
#   unnest(cols = c(metrics))
# well_summaries

Grouping

add_groupings() takes one or more strings with grouping shorthand, expands them and joins them onto the provided data. The source well(s)->target well(s). Sets of wells are specified with the familiar rectanuglar colon style notation: A01:A12 is expanded to contain row A wells from column 1 to column 12 inclusive.

# well_summaries %>% 
#   add_groupings(c("A01->A02:A12", "B01:H01->B02:H12"), nrow = 8, ncol = 12)

A more complex example (using the internal fxn for demonstration of the parsing that goes on under the hood of add_groupings())

# growr:::parse_grouping('p1!A01,B01:B04,H03->A01:H12', 8 , 12)

Relatively

Rescale you data by group, for example:

# well_summaries %>% 
#   add_groupings(c("A01->A02:A12", "B01:H01->B02:H12"), nrow = 8, ncol = 12) %>% 
#   group_by(group) %>% 
#   mutate(relative_mu = relatively(mu, is_ref)) 

---

Installation

# install.packages("growr")
remotes::install_github("npjc/growr")

---

Model Extension Interface:

growr provides many model choices, but you may want to add your own. For this reason there is a standard way to add a new model interface

new_model_fxn <- function(x, y, ...) {
  model_obj <- fit_new_model(x, y, ...)


  output <- list(model = model_summary,
                 component = component_summary,
                 observation = observation_summary
                 )
}

npjc/growr documentation built on Nov. 9, 2019, 7:29 a.m.