Adjusting the data structure | Too few time points / observations | Decreasing trend

Introduction | Fitting the models to the data | Model fitting components (h0 free) | Model parameters

Additional parameters for the sigmoidal model | Additional parameters for the double-sigmoidal model

Fitting the models to the data

The decision process

Example fit on simulated input data | The fit object

Introduction | biwt package functions | Functions for resistant estimation | biwt.est() | biwt.est(x, r = 0.2, med.init = covMcd(x)) | biwt_est() | biwt_est(x, r, med.init) | Functions for resistant correlation | biwt.cor() | biwt.cor(x, r = 0.2, output = "matrix", median = TRUE, full.init = TRUE, absval = TRUE) | biwt_cor() | biwt_cor(x, r, median = TRUE, full.init = TRUE) | biwt_cor_matrix() | biwt_cor_matrix(x, r, median = TRUE, full.init = TRUE) | biwt_dist_matrix() | biwt_dist_matrix(x, r, median = TRUE, full.init = TRUE, absval = TRUE) | Correlation | Biweight Correlation | Pearson Correlation | Comparison of Biweight and Pearson correlation on clean data | Comparison of Biweight and Pearson correlation on contaminated data | Simulation Study | Example #1: breakdown > contamination, 20% compressed contamination | Example #2: breakdown > contamination, 20% diffuse contamination | Example #3: breakdown < contamination, 40% compressed contamination | Example #4: low population correlation parameter of 0.1, 20% compressed contamination | Presidential Candidate Voting Data | Dendrograms | Biweight dendrogram | Euclidean dendrogram | Pearson dendrogram | Heatmaps | Biweight Heatmap | Euclidean Heatmap | Pearson Heatmap