[view raw Rmd]

This is the fourth blog post on quantities, an R-Consortium funded project for quantity calculus with R. It is aimed at providing integration of the ‘units’ and ‘errors’ packages for a complete quantity calculus system for R vectors, matrices and arrays, with automatic propagation, conversion, derivation and simplification of magnitudes and uncertainties. This article summarises the latest enhancements and investigates how to fit linear regressions with quantities. In previous articles, we discussed a first working prototype, units and errors parsing, and data wrangling operations with quantities.

Latest enhancements

In the following, we briefly describe some important enhancements made to the units, errors and quantities packages. Also, we would like to note that, thanks to Katharine Mullen’s careful review, packages units, errors and constants are now listed in the ChemPhys CRAN Task View.

Mixed units

Apart from various minor improvements and bug fixes, the most notable new feature is the support for mixed units, that will be released on CRAN, foreseeably, within a month.

One of the most prominent design decisions made in the units package (which applies to errors and quantities as well), following R’s philosophy, is that units objects are fundamentally vectors. This means that a units (errors, quantities) object represents one or more measurement values of the same quantity, with the same unit (for instance, repeated measurements of the same quantity). Thus, different quantities, with different units, must belong to different objects.

However, Bill Denney raised an interesting use case (#134, #145) in which different quantities need to be manipulated in a single data structure. Very briefly, he receives heterogeneous measurements of different analytes from clinical studies as follows:

(analytes <- data.frame(
  analyte=c("glucose", "insulin", "glucagon"),
  original_unit=c("mg/dL", "IU/L", "mmol/L"),
  original_value=c(1, 2, 3),
  new_unit=c("mmol/L", "mg/dL", "mg/L"),
  stringsAsFactors=FALSE
))
##    analyte original_unit original_value new_unit
## 1  glucose         mg/dL              1   mmol/L
## 2  insulin          IU/L              2    mg/dL
## 3 glucagon        mmol/L              3     mg/L

To be able to convert these values to the new units, first we need to define some conversion constants between grams and IUs (which stands for International Unit) or moles of a particular substance (note: numbers may be wrong):

# some adjustments
(analytes <- within(analytes, {
  for (i in seq_along(analyte)) {
    original_unit[i] <- gsub("(mol|IU)", paste0("\\1_", analyte[i]), original_unit[i])
    new_unit[i] <- gsub("(mol|IU)", paste0("\\1_", analyte[i]), new_unit[i])
  }
  i <- NULL
}))
##    analyte   original_unit original_value       new_unit
## 1  glucose           mg/dL              1 mmol_glucose/L
## 2  insulin    IU_insulin/L              2          mg/dL
## 3 glucagon mmol_glucagon/L              3           mg/L
library(units)
## udunits system database from /usr/share/udunits
install_conversion_constant("mol_glucose", "g", 180.156)
install_conversion_constant("g", "IU_insulin", 25113.32)
install_conversion_constant("mol_glucagon", "g", 3482.80)

Then, the development version of units provides a new method called mixed_units() intended for this use case:

(analytes <- within(analytes, {
  original_value <- mixed_units(original_value, original_unit)
  new_value <- set_units(original_value, new_unit)
  original_unit <- new_unit <- NULL
}))
##    analyte      original_value                   new_value
## 1  glucose           1 [mg/dL] 0.05550745 [mmol_glucose/L]
## 2  insulin    2 [IU_insulin/L]         0.007963901 [mg/dL]
## 3 glucagon 3 [mmol_glucagon/L]              10448.4 [mg/L]

Mixed units are basically lists with a custom class, and each element of the list is a units object:

analytes$original_value
## Mixed units: IU_insulin/L (1), mg/dL (1), mmol_glucagon/L (1)
## 1 [mg/dL], 2 [IU_insulin/L], 3 [mmol_glucagon/L]
class(analytes$original_value)
## [1] "mixed_units"
unclass(analytes$original_value)
## [[1]]
## 1 [mg/dL]
##
## [[2]]
## 2 [IU_insulin/L]
##
## [[3]]
## 3 [mmol_glucagon/L]
class(analytes$original_value[[1]])
## [1] "units"

Still, units objects cannot be concatenated into mixed lists unless explicitly enabled by the user, thus maintaining backwards compatibility:

c(as_units("m"), as_units("s")) # error, cannot convert, cannot mix
## Error in c.units(as_units("m"), as_units("s")): units are not convertible, and cannot be mixed; try setting units_options(allow_mixed = TRUE)?
c(as_units("m"), as_units("s"), allow_mixed=TRUE)
## Mixed units: m (1), s (1)
## 1 [m], 1 [s]

This behaviour can be controlled also by the global option allow_mixed (see help(units_options)). Finally, note that mixed units with non-heterogeneous units are not simplified either unless explicitly requested:

(x <- mixed_units(1:3, c("m", "s", "m")))
## Mixed units: m (2), s (1)
## 1 [m], 2 [s], 3 [m]
as_units(x) # error, cannot convert, cannot mix
## Error in c.units(structure(1L, units = structure(list(numerator = "m", : units are not convertible, and cannot be mixed; try setting units_options(allow_mixed = TRUE)?
x[c(1, 3)]
## Mixed units: m (2)
## 1 [m], 3 [m]
as_units(x[c(1, 3)])
## Units: [m]
## [1] 1 3

Compatibility with this feature has been also added to the quantities package. Specifically, lists of mixed units can contain either units or quantities objects, and additional methods have been defined to deal with them transparently.

library(quantities)
## Loading required package: errors
c(set_quantities(1, m, 0.1), set_quantities(2, s, 0.3), allow_mixed=TRUE)
## Mixed units: m (1), s (1)
## 1.0(1) [m], 2.0(3) [s]
(x <- mixed_units(set_errors(1:2, c(0.1, 0.3)), c("m", "km")))
## Mixed units: km (1), m (1)
## 1.0(1) [m], 2.0(3) [km]
as_units(x)
## Units: [m]
## Errors:   0.1 300.0
## [1]    1 2000
# etc.

Of course, parsers also aware of this new feature (see also the new vignette on parsing quantities):

parse_quantities(c("1.02(5) g", "2.51(0.01) V", "(3.23 +/- 0.12) m"))
## Mixed units: g (1), m (1), V (1)
## 1.02(5) [g], 2.51(1) [V], 3.2(1) [m]

We kindly invite the community to try out this new feature (currently on GitHub only) and report any issue or proposal for improvement.

Support for correlations

Version 0.3.0 of errors hit CRAN a month ago with a very important feature that was missing before: support for correlations between quantities.

Due to the design of these packages, as discussed before, the advantage of having separate vectorised variables to operate freely with them without having to build an expression (as in the propagate package, for example) makes it harder to store pairwise correlations and operate with them. This has been finally resolved in this version thanks to an internal hash table, which automatically cleans up dangling correlations when the associated objects are garbage-collected.

The manual page help("errors-package") provides a nice introductory example on how to set up correlations and how these are propagated (see help("correl") for more detailed information):

library(errors)

# Simultaneous measurements of voltage, intensity and phase
GUM.H.2
##       V        I    phi
## 1 5.007 0.019663 1.0456
## 2 4.994 0.019639 1.0438
## 3 5.005 0.019640 1.0468
## 4 4.990 0.019685 1.0428
## 5 4.999 0.019678 1.0433
# Obtain mean values and uncertainty from measured values
V   <- mean(set_errors(GUM.H.2$V))
I   <- mean(set_errors(GUM.H.2$I))
phi <- mean(set_errors(GUM.H.2$phi))

# Set correlations between variables
correl(V, I)   <- with(GUM.H.2, cor(V, I))
correl(V, phi) <- with(GUM.H.2, cor(V, phi))
correl(I, phi) <- with(GUM.H.2, cor(I, phi))

# Computation of resistance, reactance and impedance values
(R <- (V / I) * cos(phi))
## 127.73(7)
(X <- (V / I) * sin(phi))
## 219.8(3)
(Z <- (V / I))
## 254.3(2)
# Correlations between derived quantities
correl(R, X)
## [1] -0.5884298
correl(R, Z)
## [1] -0.4852592
correl(X, Z)
## [1] 0.9925116

In a similar way, correlations transparently work with quantities objects. For example, let us suppose that we measured the position of a particle at several time instants:

library(quantities)

x <- set_quantities(1:5, m, 0.05)
t <- set_quantities(1:5, s, 0.05)

Each measurement has some uncertainty (the same for all values here for simplicity). Now we can compute the distance covered in each interval, and then the instantaneous velocity, which is constant here:

dx <- diff(x)
dt <- diff(t)
(v <- dx/dt)
## Units: [m/s]
## Errors: 0.1 0.1 0.1 0.1
## [1] 1 1 1 1

Obviously, there should be a strong correlation between the instantaneous velocity and the distance covered for each interval. And here it is:

correl(dx, v)
## [1] 0.7071068 0.7071068 0.7071068 0.7071068

Fitting linear models with quantities

A linear regression models the relationship between a dependent variable and one or more explanatory variables. These variables are usually quantities, some measurements with some unit and uncertainty associated. Therefore, the output from a linear regression (coefficients, fitted values, predictions…) are quantities as well. However, functions such as lm are not compatible with quantities. This section describes current issues and discusses several approaches to overcome them, along with their benefits, advantages and limitations.

Current issues

Let us generate some artificial data with the classical formula for uniformly accelerated movement, \(s(t) = s_0 + v_0t + \frac{1}{2}at^2\):

library(quantities)
set.seed(1234)

t <- seq(1, 10, 0.1)
s <- 3 + 2*t + t^2

# some noise added
df <- data.frame(
  t = set_quantities(t + rnorm(length(t), 0, 0.01), s, 0.01),
  s = set_quantities(s + rnorm(length(t), 0, 1), m, 1)
)
plot(df)

Then, we try to adjust a linear model using lm:

fit <- lm(s ~ poly(t, 2), df) # error Ops.units
## Error in Ops.units(X, Y, ...): power operation only allowed with length-one numeric power

First issue: it seems that poly computes powers in a vectorised way (i.e., t^0L:degree), which is not currently supported in units, because it would generate different units for each value. Now that mixed units are supported, this could be a way to circumvent this, but it is not clear whether the resulting list of mixed units may create more problems. This is something that we should explore anyway.

Let us try this time by explicitly defining the powers:

(fit <- lm(s ~ t + I(t^2), df))
## Warning: In 'Ops' : non-'errors' operand automatically coerced to an
## 'errors' object with no uncertainty

##
## Call:
## lm(formula = s ~ t + I(t^2), data = df)
##
## Coefficients:
## (Intercept)            t       I(t^2)
##       3.373        1.910        1.006

Now it works. We obtain the (unitless, errorless) coefficients, and these are other parameters and summaries:

coef(fit) # plain numeric, as show above
## (Intercept)           t      I(t^2)
##    3.373459    1.909901    1.006140
residuals(fit)[1:5] # wrong uncertainty, copied from 's'
## Units: [m]
## Errors: 1 1 1 1 1
##           1           2           3           4           5
##  0.01289180  1.41273622  0.68031389 -0.65668199  0.07589058
fitted(fit)[1:5] # wrong uncertainty
## Units: [m]
## Errors: 0 0 0 0 0
##        1        2        3        4        5
## 6.242304 6.703228 7.161199 7.451099 8.039660
predict(fit, data.frame(t=11:15)) # plain numeric
##        1        2        3        4        5
## 146.1253 171.1765 198.2399 227.3156 258.4035
summary(fit) # error Ops.units
## Error in Ops.units(mean(f)^2, var(f)): both operands of the expression should be "units" objects

In summary, we do not get the benefit of obtaining coefficients, fitted values, predictions… with the right units and uncertainty, and the whole object is a mess due to diverse incompatibilities.

Wrapping linear models

There are several possible ways to overcome the issues above. The most direct one would be to wrap the lm call, so that quantities are dropped before calling lm, and the resulting object is modified to set up the proper quantities a posteriori. However, in this way, some lm methods may work while some others may still be broken.

A cleaner approach would be to wrap the lm call to add a custom class to the hierarchy and save units and errors for later use:

qlm <- function(formula, data, ...) {
  # get units info, then drop quantities
  row <- data[1,]
  for (var in colnames(data)) if (inherits(data[[var]], "quantities")) {
    data[[var]] <- drop_quantities(data[[var]])
  }

  # fit linear model and add units info for later use
  fit <- lm(formula, data, ...)
  fit$units <- lapply(eval(attr(fit$terms, "variables"), row), units)
  class(fit) <- c("qlm", class(fit))
  fit
}

(fit <- qlm(s ~ t + I(t^2), df))
##
## Call:
## lm(formula = formula, data = data)
##
## Coefficients:
## (Intercept)            t       I(t^2)
##       3.373        1.910        1.006
class(fit)
## [1] "qlm" "lm"

Then, this custom class can be used to build specific methods of interest:

coef.qlm <- function(object, ...) {
  # compute coefficients' units
  coef.units <- lapply(object$units, as_units)
  for (i in seq_len(length(coef.units)-1)+1)
    coef.units[[i]] <- coef.units[[1]]/coef.units[[i]]
  coef.units <- lapply(coef.units, units)

  # use units above and vcov diagonal to set quantities
  coef <- mapply(set_quantities, NextMethod(), coef.units,
                 sqrt(diag(vcov(object))), mode="symbolic", SIMPLIFY=FALSE)

  # use the rest of the vcov to set correlations
  p <- combn(names(coef), 2)
  for (i in seq_len(ncol(p)))
    covar(coef[[p[1, i]]], coef[[p[2, i]]]) <- vcov(fit)[p[1, i], p[2, i]]

  coef
}

coef(fit)
## $`(Intercept)`
## 3.4(5) [m]
##
## $t
## 1.9(2) [m/s]
##
## $`I(t^2)`
## 1.01(2) [m/s^2]
fitted.qlm <- function(object, ...) {
  # set residuals as std. errors of fitted values
  set_quantities(NextMethod(), object$units[[1]],
                 residuals(object), mode="symbolic")
}

fitted(fit)[1:5]
## Units: [m]
## Errors: 0.01289180 1.41273622 0.68031389 0.65668199 0.07589058
##        1        2        3        4        5
## 6.242304 6.703228 7.161199 7.451099 8.039660
predict.qlm <- function(object, ...) {
  # set se.fit as std. errors of predictions
  set_quantities(NextMethod(), object$units[[1]],
                 NextMethod(se.fit=TRUE)$se.fit, mode="symbolic")
}

predict(fit, data.frame(t=11:15))
## Units: [m]
## Errors: 0.4570381 0.6507279 0.8804014 1.1448265 1.4434174
##        1        2        3        4        5
## 146.1253 171.1765 198.2399 227.3156 258.4035

and so on and so forth.

Open problems

This analysis is limited to the lm function, but there are others, both in R base (such as glm) and in other packages, which have different sets of input parameters and output. Instead of developing multiple sets of wrappers and methods, it would be desirable to manage everything through a common wrapper, class and set of methods (see, e.g., how ggplot2::geom_smooth works). It should be assessed whether this is possible, at least for a limited, widely-used, group of functions.

Also, there may be users interested in fitting linear models with units only, or with uncertainty only. As with the rest of the functionalities in these packages, it should be studied how to wisely break down this feature.

Summary

This article summarises the latest enhancements in the units, errors and quantities packages, and provides some initial prospects on fitting linear models with quantities. Also, this is the last deliverable of the R-quantities project, which has reached the following milestones:

  1. A first working prototype.
  2. Support for units and errors parsing.
  3. An analysis of data wrangling operations with quantities.
  4. Prospects on fitting linear models with quantities.

And along the way, there have been multiple exciting improvements, both in units and errors, to support all these features and make quantities possible, which is ready for an imminent CRAN release. This project ends, but the r-quantities GitHub organisation will continue to thrive and to provide the best tools for quantity calculus to the R community.

Acknowledgements

This project gratefully acknowledges financial support from the R Consortium. Also I would like to thank Edzer Pebesma for his continued support and collaboration.