The following examples demonstrate the use of IRootLab library to analyse data and generate figures. The directories are roughly organized by dataset.
DATA_ARTIFICIAL | 2-feature dataset created using gendata. Most demos here show 2-dimensional classification regions |
| DEMO_BAGGING_SVM | Bagging example using SVM classifier and drawing classification regions. \|/ → |
| DEMO_CLSSR_D | Draws classification regions for classifiers LDC/QDC \|/ → |
| DEMO_CLSSR_KNN | Draws classification regions for classifier k-NN \|/ → |
| DEMO_CLSSR_TREE | Draws classification regions for the Tree classifier \|/ → |
| DEMO_ECLASS_ANIMATION | Shows evolution of classifier eClass, saves animated GIF. \|/ → |
| DEMO_ECLASS_ARTIFICIAL | Draws classification regions for classifier eClass \|/ → |
DATA_BRAIN | Ketan's brain cancer dataset, no=439, nf=235, 3 classes |
| DEMO_RATER_BRAIN | Classification of Brain data using LDC classifier, leave-one-out cross-validation \|/ → |
| DEMO_U_TEST_PER_WAVENUMBER | U-test per wavenumber is performed for one-versus-one datasets (2-class datasets) \|/ → |
DATA_SHE | Syrian Hamster Embryo (SHE) assay dataset, 2 class levels: <chemical>|<transformation>; 5 * 2 = 10 classes; no=600; nf=58 |
BOTH | Use both levels of SHE dataset in the analysis |
| DEMO_BMTABLE | Biomarkers of Non-transformed vs. Transformed, separated by Chemical \|/ → |
| DEMO_DECIDERS_AND_GRAGS | Different ways to use per-spectrum prediction aggregation (grag), and decision thresholds (decider) \|/ → |
CHEMICALS | Analyses using class level 1 only: {'B[a]P', '3-MCA', 'Ant', '2,4dT', 'o-T'} |
| DEMO_CROSSCALCULATED_LDA | Cross-calculated LDA scores \|/ → |
| DEMO_PCALDA | PCA-LDA demo, scores plots, cluster vectors \|/ → |
| DEMO_RATER | Classification of Chemicals using LDC and cross-validation \|/ → |
TRANSFORMATION | Analyses using class level 2 only: {'Non-transformed', 'Transformed'} |
| DEMO_AS_DSPERC_X_RATE | (dataset %) x (classification rate %) curve to check sample size \|/ → |
| DEMO_DOES_BAGGING_HELP | Tracking the improvement of classification with addition of component classifiers \|/ → |
| DEMO_FACTORSCURVE | (number of factors)X(Classification rate) curve \|/ → |
| DEMO_FEATURE_HISTOGRAM_COLORS | Shows different ways to paint the same Feature Histogram \|/ → |
| DEMO_FORWARD | Forward feature selection demo \|/ → |
| DEMO_GRIDSEARCH_KNN_AND_PCA | Combined optimization of PCA number of factors & k-NN k \|/ → |
| DEMO_GRIDSEARCH_KNN_K | Grid search to obtain best k-NN's k \|/ → |
| DEMO_GRIDSEARCH_PCA_DISCRIMINANT | Grid search to simultaneously optimize (PCA number of factors) x ('linear'/'quadratic') \|/ → |
DATA_WINE | Wine dataset from UCI repository - http://archive.ics.uci.edu/ml/datasets/Wine; no=178; nf=13; nc=3 |
| DEMO_ECLASS_INCREMENTAL_LEARNING | Increase of classification rate as eClass is incrementally trained \|/ → |
| DEMO_SVM_C_GAMMA | Grid search optimization of SVM (C, gamma) (Gaussian Kernel) \|/ → |
INTROSPECTION | Self-checking of IRootLab directory |
| DEMO_CLASSES_HTML | Generates IRootLab classes hierarchical tree in HTML, using object colors. \|/ → |
| DEMO_CLASSES_TXT | Creates a CSV file with a list of classes, similar to what is seen in blockmenu.m \|/ → |
OTHER_DATA | Others and other data |
| DEMO_IMPORT_FISHERIRIS | Shows how to assemble a dataset from existing MATLAB matrices (Fisher Iris data example) \|/ → |
PREPROCESSING | Pre-processing demos |
| DEMO_PRE_BC_RUBBER | Demonstrates the Convex Polynomial Line baseline correction \|/ → |
| DEMO_RAMAN_PREPROCESS | Pre-processing of Raman data: Wavelet-De-noising, Polynomial Baseline Correction, Vector Normalization \|/ → |
SAMPLEDATA | Datasets included with IRootLab releases. After clicking on file, new dataset can be handled in objtool |
| IROOTLAB_SETUP | % V V \|/ → |
| LOAD_DATA_HINT | Loads the hint dataset: this dataset containg one spectrum only: 1800-900 cm^-1 \|/ → |
| LOAD_DATA_KETAN_BRAIN_ATR | Loads Ketan's brain cancer dataset \|/ → |
| LOAD_DATA_MATT_NANOPARTICLES_SYNCHROTRON | Loads Matt's synchrotron data (5 spectra only) \|/ → |
| LOAD_DATA_RAMAN_SAMPLE | Loads sample data raman_sample.mat \|/ → |
| LOAD_DATA_SHE5TRAYS | Loads sample data she5trays.mat \|/ → |
| LOAD_DATA_UCI_WINE | Loads sample data userdata_nc2nf2.txt \|/ → |
| LOAD_DATA_UGLYSPECTRUM | Loads sample data uglyspectrum.mat \|/ → |
| LOAD_DATA_USERDATA_NC2NF2 | Loads sample data userdata_nc2nf2.txt \|/ → |
| SAMPLEDATA_VIEW_ALL | Plots all sample datasets in separate figures \|/ → |