Skip to contents

measure

Source: vignettes/measure.Rmd
measure.Rmd
library(measure)
#> Loading required package: recipes
#> Loading required package: dplyr
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
#> 
#> Attaching package: 'recipes'
#> The following object is masked from 'package:stats':
#> 
#>     step

To understand measure, we need to re-introduce some concepts from recipes, namely variables and roles. Variables refer the the data columnn in a data frame or tibble that contain raw data. Roles define how we will use these variables in our model. For analytical measurements, we need more than the roles that recipes provides. Recipes was designed to be extensible in exactly this way.

Starting from Max Kuhn’s suggestions, additional roles to be included in this package are:

  • measurment variables: raw data from an instrument (e.g., time, absorbance, millivolts, volume, etc.)
  • sample identifies: define the subset of the data that we are interested in (e.g., sample, condition, replicate, day, etc.)
  • experimental conditions: define the conditions under which the measurement was made (e.g., temperature, pH, reagents, column type, flow rate, etc.)

Our challenge is to strike the right balance between a flexibile package and a useful one.

Scope of Data Sources

Analytical measurement systems are as varied as the applications they are used in. The field is necessarily broad and growing. The American Chemical Societydefines the field as:

the science of obtaining, processing, and communicating information about the composition and structure of matter. In other words, it is the art and science of determining what matter is and how much of it exists.

Common analytical techniques include chromatography, optical spectroscopy, mass spectrometry, microscopy, and many others. The initial scope of measure is on common techniques that can represent data in a tibble or series of tibbles.

Researchers who are experts in performing these types of measurements refer their equipment as instruments. However, instrumental variables already have a meaning in the statistics literature, and so we will avoid that terminology.

Data Sources

There are a number of data sets in modeldata and prospectr packages. These include modeldata::meats and prospectr::NIRsoil.

Mendeley Data offers data inlcuded as part of various publications. A few possibilities include:

There are also some repositories that might be of use: