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:
Our challenge is to strike the right balance between a flexibile package and a useful one.
Analytical measurement systems are as varied as the applications they are used in. The field is necessarily broad and growing. The American Chemical Society defines 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.
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: