gasfluxes: Flux calculation

Description

A wrapper function for convenient flux calculation.

Usage

gasfluxes(dat, .id = "ID", .V = "V", .A = "A", .times = "time",
  .C = "C", methods = c("linear", "robust linear", "HMR", "NDFE"),
  k_HMR = log(1.5), k_NDFE = log(0.01), verbose = TRUE,
  plot = TRUE, select, maxiter = 100, ...)

Arguments

dat

a data.frame or data.table with data from flux measurements.

.id

character vector specifying the columns to be used as ID, multiple ID columns are possible.

character specifying the column containing chamber volume values.

character specifying the column containing chamber area values.

.times

character specifying the column containing chamber closing time values.

character specifying the column containing concentration values.

methods

character; which methods to use for flux estimation. See details for available methods.

k_HMR

starting value for HMR.fit.

k_NDFE

starting value for NDFE.fit.

verbose

logical; print progress messages?

plot

create plots if TRUE (the default). The IDs are used as file names and should thus not include characters that are not allowed in file names, such as / or =. A directory "pics" is created in the working directory if it doesn't exist. The plots are only intended to facilitate quick checking, not for publication quality graphs.

select

deprecated; please use function selectfluxes.

maxiter

see nls.control

…

further parameters

Value

A data.table with the results of the flux calculation. See the documentation of the fitting functions for details. If a selection algorithm has been specified, the last columns are the selected flux estimate, the corresponding standard error and p-value and the method with which the selected flux was estimated.

Details

Available methods are

"linear":	`lin.fit`
"robust linear":	`rlin.fit`
"HMR":	`HMR.fit`
"original HMR":	`HMR.orig`
"NDFE":	`NDFE.fit`

Specifying other methods results in an error.

The default starting values for "HMR" and "NDFE", \(k = log(\kappa)\) and \(k = log(\tau)\), resp., assume that time is in hours. If you use a different time unit, you should adjust them accordingly. Note that nls is used internally by these functions and thus they should not be used with artificial "zero-residual" data.

The input data.frame or data.table should be in the following format:

   serie        V A      time         C
1:    ID1 0.522625 1 0.0000000 0.3317823
2:    ID1 0.522625 1 0.3333333 0.3304053
3:    ID1 0.522625 1 0.6666667 0.3394311
4:    ID1 0.522625 1 1.0000000 0.4469102
5:    ID2 0.523625 1 0.0000000 0.4572708

However, more than one ID column are possible. E.g., the first ID column could be the plot and a second ID column could be the date. Keep in mind that the combination of IDs must be a unique identifier for each flux measurement.

Units of the output depend on input units. It's recommended to use [V] = m^3, [A] = m^2, [time] = h, [C] = [mass or mol]/m^3, which results in [f0] = [mass or mol]/m^2/h. Since all algorithms use V/A, A can be input as 1 and V as the chamber height.

Examples

Run this code

# NOT RUN {
#compare result of original HMR with plinear HMR
data(fluxMeas)
res <- gasfluxes(fluxMeas[1:400,], 
                 .id = "serie", .V = "V", .A = "A",
                 .times = "time", .C = "C",
                 methods = c("HMR", "original HMR"), verbose = TRUE)
                 
#number of successful fits
res[, sum(!is.na(HMR.kappa))]
res[, sum(!is.na(original.HMR.kappa))]

#Do the results differ?
plot(res[["HMR.f0"]], res[["original.HMR.f0"]])
abline(0, 1)


res <- gasfluxes(fluxMeas, 
                 .id = "serie", .V = "V", .A = "A",
                 .times = "time", .C = "C",
                 methods = "HMR", verbose = TRUE)
# Error: time not sorted in flux ID ID556. 
# Investigate the problem:
fluxMeas[serie %in% c("ID555", "ID556", "ID557")]
#    serie        V A      time         C
# 1: ID555 0.551625 1 0.0000000 0.3884388
# 2: ID555 0.551625 1 0.3333333 0.4125270
# 3: ID555 0.551625 1 0.6666667 0.3714207
# 4: ID555 0.551625 1 1.0000000 0.3735092
# 5: ID556 0.524250 1 0.0000000 0.3638239
# 6: ID556 0.524250 1 0.3333333 0.3520481
# 7: ID556 0.524250 1 0.6666667 0.3551644
# 8: ID557 0.528375 1 0.0500000 0.3954601
# 9: ID556 0.524250 1 0.0000000 0.3839834
#10: ID557 0.528375 1 0.3333333 0.3967269
#11: ID557 0.528375 1 0.6666667 0.3764967
#12: ID557 0.528375 1 1.0000000 0.3973055

# some mixup of IDs and times
# usually an input or Excel error during data preparation
# investigate and fix
# }
# NOT RUN {
# }