sicegar (version 0.3.0)

categorize_h0: Categorize input data by comparing the AIC values of the three fitted models.

Description

Categorizes the input data using the results of two model fits and chosen thresholds.

Usage

categorize_h0(
  parameterVectorSigmoidal,
  parameterVectorDoubleSigmoidal,
  threshold_intensity_range = 0.1,
  threshold_minimum_for_intensity_maximum = 0.3,
  threshold_bonus_sigmoidal_AIC = 0,
  threshold_sm_tmax_IntensityRatio = 0.85,
  threshold_dsm_tmax_IntensityRatio = 0.75,
  threshold_AIC = -10,
  threshold_t0_max_int = 1e+10,
  showDetails = FALSE
)

Value

The returned object contains extensive information about the decision process, but the key component is the decision variable. The decision variable can be one of the following four; "no_signal", "sigmoidal","double sigmoidal" or "ambiguous".

Arguments

parameterVectorSigmoidal

Output of the sigmoidalFitFunction.

parameterVectorDoubleSigmoidal

Output of the doublesigmoidalFitFunction.

threshold_intensity_range

Minimum for intensity range, i.e. it is the lower limit for the allowed difference between the maximum and minimum of the intensities (Default is 0.1, and the values are based on actual, not the rescaled data.).

threshold_minimum_for_intensity_maximum

Minimum allowed value for intensity maximum. (Default is 0.3, and the values are based on actual, not the normalized data.).

threshold_bonus_sigmoidal_AIC

Bonus AIC points for sigmoidal fit. Negative values help the sigmoidal model to win. Only helps in competition between sigmoidal and double sigmoidal fit at decision step "9", i.e. if none of the models fail in any of the tests and stay as a candidate until the last step (Default is 0).

threshold_sm_tmax_IntensityRatio

The threshold for the minimum intensity ratio between the intensity at the last observed time points and the theoretical maximum intensity of the sigmoidal curve. If the value is below the threshold, then the data can not be represented with the sigmoidal model. (Default is 0.85)

threshold_dsm_tmax_IntensityRatio

The threshold for the minimum intensity ratio between the intensity at the last observed time points and the maximum intensity of the double sigmoidal curve. If the value is above the threshold, then the data can not be represented with the double sigmoidal model. (Default is 0.75)

threshold_AIC

Maximum AIC values in order to have a meaningful fit (Default is -10).

threshold_t0_max_int

Maximum allowed intensity of the fitted curve at time is equal to zero (t=0). (Default is 1E10, and the values are based on actual, not the normalized data.).

showDetails

Logical to choose if we want to see details or not. Default is "FALSE"

Examples

# runif() is used here for consistency with previous versions of the sicegar package. However,
# rnorm() will generate symmetric errors, producing less biased numerical parameter estimates.
# We recommend errors generated with rnorm() for any simulation studies on sicegar.
# Example 1 with double sigmoidal data
time=seq(3, 24, 0.1)

#simulate intensity data and add noise
noise_parameter <- 0.2
intensity_noise <- runif(n = length(time), min = 0, max = 1) * noise_parameter
intensity <- sicegar::doublesigmoidalFitFormula_h0(time,
                                               finalAsymptoteIntensityRatio = .3,
                                               maximum = 4,
                                               slope1Param = 1,
                                               midPoint1Param = 7,
                                               slope2Param = 1,
                                               midPointDistanceParam = 8,
                                               h0 = 1)
intensity <- intensity + intensity_noise

dataInput <- data.frame(intensity = intensity, time = time)
normalizedInput <- sicegar::normalizeData(dataInput,
                                         dataInputName = "sample001")


# Fit sigmoidal model
sigmoidalModel <- sicegar::multipleFitFunction_h0(dataInput = normalizedInput,
                                              model = "sigmoidal",
                                              n_runs_min = 20,
                                              n_runs_max = 500,
                                              showDetails = FALSE)

# Fit double sigmoidal model
doubleSigmoidalModel <- sicegar::multipleFitFunction_h0(dataInput = normalizedInput,
                                                    model = "doublesigmoidal",
                                                    n_runs_min = 20,
                                                    n_runs_max = 500,
                                                    showDetails = FALSE)


# Calculate additional parameters
sigmoidalModel <- sicegar::parameterCalculation_h0(sigmoidalModel)
doubleSigmoidalModel <- sicegar::parameterCalculation_h0(doubleSigmoidalModel)

outputCluster <- sicegar::categorize_h0(parameterVectorSigmoidal = sigmoidalModel,
                                    parameterVectorDoubleSigmoidal = doubleSigmoidalModel)

utils::str(outputCluster)