Generate synthetic mixtures from pure spectra
A Python script to generate random synthetic mixtures from pure spectra from a database. This post can be used as a tutorial to generate datasets for Multivariate Curve Resolution.
Diagnostic plots for linear regression
A short survey of diagnostic plots, as tools to dig deeper on the assumptions behind a regression model.
Genetic algorithm for wavelength selection using Numpy
Regression, Partial Least Squares Regression, Variable Selection
06/22/2024
A implementation of a genetic algorithm for wavelength selection using basic Numpy functions.
Robust PCA
A implementation of robust PCA, useful when the data contains outliers.
Two methods for baseline correction of spectral data
Worked examples of two methods for baseline correction of spectra applied to Raman and XRF data.
Optimal spectra smoothing with Fourier ring correlation
We bring in the concept of Fourier ring correlation for another riff on the optimal settings of smoothing filters.
Spectra smoothing with locally-weighted regression
Locally weighted regression is an approach for smoothing a set of data points by performing a polynomial regression in the vicinity of each point. It does sounds like a …
Wavelength selection with a genetic algorithm
Wavelength selection methods aim at choosing the spectral bands that produce the best regression or classification model. Here we introduce a genetic algorithm for wavelength selection.
Multi-class classification for NIR spectroscopy: the very basics
Multi-class classification aims and subdividing samples into one of multiple predefined categories. In this post we explore a basic classifier and discuss important metrics such as accuracy and AUC …
Updates and additions to the PLS Regression code
Regression, Partial Least Squares Regression, Plots and Charts
04/29/2023
Updated code and additional utility scripts for PLS regression. Will keep it updated as we go.
Multivariate curve resolution: an introduction
Multivariate Curve Resolution deals with spectra, or other signals, from samples containing multiple components, and aims at recovering the pure components.
Wavelet denoising of spectra
On the concept of wavelet transform and decomposition, and its application for denoising of spectra.
Understanding neural network parameters with TensorFlow in Python: the optimiser
Neural Networks, Regression
11/20/2022
An introductory tutorial on optimisers for deep learning, including Python code for a regression training for NIR spectroscopy.
The Kennard-Stone algorithm
Data Operations and Plotting
10/15/2022
The Kennard-Stone algorithm is a criterion to divide a dataset into a calibration and a validation subset. Here's an introduction to the method and a basic Python implementation.
Detecting lactose in lactose-free milk with NIR spectroscopy
Use Cases
09/24/2022
NIR spectroscopy can detect sources of allergens or intolerance in food. Here's an example of detecting lactose in lactose-free milk.
The PCA correlation circle
The PCA correlation circle is a useful tool to visually display the correlation between spectral bands and principal components. The correlation can be quantified through the Euclidean distance and …
Regression optimisation with a Pipeline
The process of developing and optimising a regression model, almost invariably requires a sequence of steps. These steps can be combined in a single predictor using the Pipeline function …
Qualitative analysis of ground coffee with NIR spectroscopy
Use Cases
07/30/2022
Can we use NIR spectroscopy to distinguish between coffee types? Let's take a look at how basic PCA can enable a qualitative estimation of coffee strength.
Aquagrams with Python and Matplotlib
Data Operations and Plotting, Plots and Charts
07/09/2022
Aquagrams are diagrams charting the relative strength of the NIR reflectance or absorbance of specific bands in water-based systems. Here's an example code to plot an aquagram using Matplotlib.
Parallel computation of loops for cross-validation analysis
Using parallel computation to speed up cross-validation analysis for large data sets.
Understanding neural network parameters with TensorFlow in Python: the activation function
Neural Networks, Regression
04/09/2022
Where we discuss the meaning of an activation function in neural networks, discuss a few examples, and show a comparison of neural network training with different activation functions.
Deep neural networks for spectral data regression with TensorFlow
Neural Networks, Regression
03/12/2022
This post introduces basic Python code to build fully-connected deep neural networks with TensorFlow for regression analysis of spectral data.
The Akaike Information Criterion for model selection
The Akaike Information Criterion (AIC) is another tool to compare prediction models. AIC combines model accuracy and parsimony in a single metric and can be used to evaluate data …
Minimal prediction models for linear regression
Regression
04/10/2021
What is the minimum amount of information required to export and re-use a linear regression model? The answer is surprisingly simple. Here's a step by step example using PLS …
Backward Variable Selection for PLS regression
Regression, Partial Least Squares Regression
03/13/2021
Backward Variable Selection for PLS regression is a method to discard variables that contribute poorly to the regression model. Here's a Python implementation of the method.
The Concordance Correlation Coefficient
Regression, Regression metrics, Regression Model Validation
01/09/2021
The Concordance Correlation Coefficient (CCC) can be useful to quantify the quality of a linear regression model. In this tutorial we explain the CCC and describe its relation with …
Bias-Variance trade-off in PLS regression
Bias-Variance trade-off refers to the optimal choice of parameters in a model in order to avoid both overfitting and underfitting. Let's look at a worked example using PLS regression.
Wavelength band selection with simulated annealing
Regression, Partial Least Squares Regression
08/15/2020
Improve the performance of a PLS method by wavelength band selection using Simulated Annealing optimisation.
PCA and kernel PCA explained
Principal Components Analysis
06/10/2020
Gain a practical understanding of PCA and kernel PCA by learning to code the algorithms and test it on real spectroscopic data.
Binary classification of spectra with a single perceptron
Classification, Perceptron, PLS Discriminant Analysis
04/17/2020
The perceptron is a basic block of feed-forward neural networks. Learn how to use a single perceptron for binary classification of NIR spectra using gradient descent
PLS Discriminant Analysis for binary classification in Python
Classification, PLS Discriminant Analysis
03/29/2020
PLS Discriminant analysis is a variation of PLS able to deal with classification problems. Here's a tutorial on binary classification with PLS-DA in Python
Principal component selection with simulated annealing
Principal Components Regression, Regression
02/09/2020
Simulated annealing helps overcome some of the shortcomings of greedy algorithms. Here's a tutorial on simulated annealing for principal components selection in regression.
Principal component selection with a greedy algorithm
Principal Components Regression, Regression
01/28/2020
Greedy algorithms are commonly used to optimise a function over a parameter space. Here's an implementation of a greedy algorithm for principal components selection in regression.
Choosing the optimal parameters for a Savitzky–Golay smoothing filter
Setting the parameters of a Savitzky-Golay filter seems more a craft than a science. Here's my method to find an optimal filter, complete with code.
Moving window PLS regression
Regression, Partial Least Squares Regression
12/07/2019
Not all wavelengths are created equals. A moving window PLS algorithm optimises the regression by discarding bands that are not useful for prediction.
K-fold and Montecarlo cross-validation vs Bootstrap: a primer
Cross-validation is a standard procedure to quantify the robustness of a regression model. Compare K-Fold, Montecarlo and Bootstrap methods and learn some neat trick in the process.
Fourier spectral smoothing method
Meet a fairly unknown member of the spectral smoothing family: the Fourier spectral smoothing method. Learn some theory and Python code implementation.
Savitzky–Golay smoothing method
The secret behind perfect smoothing is a wise choice of parameters. In this tutorial you will learn about the Savitzky–Golay method and the way to optimise its performance
Principal Component Regression in Python revisited
Principal Components Regression, Regression
09/10/2019
Want to get more out of your principal components regression? Here's a simple hack that will give you a stunning improvement on the performance of PCR.
Detecting outliers using the Mahalanobis distance with PCA in Python
How do we make sure we are detecting only true outliers and not cherry-picking from the data? Here's a method based on the Mahalanobis distance with PCA.
NIR data correlograms with Seaborn in Python
Data Operations and Plotting, Plots and Charts
01/03/2019
Exploratory analysis is an essential part of data analysis. Learn a handy way to explore your dataset with NIR data correlograms with Seaborn in Python.
Classification of NIR spectra by Linear Discriminant Analysis in Python
Classification, Linear Discriminant Analysis
12/03/2018
What is Linear Discriminant Analysis and how it differs from PCA? Let's talk trough LDA and build a NIR spectra classifier using LDA in Python.
Principal Components Regression vs Ridge Regression on NIR data in Python
Principal Components Regression, Regression, Ridge Regression
10/19/2018
Principal components regression is a staple of NIR analysis. Ridge regression is much used of machine learning. How do they relate? Find out in this post.
Outliers detection with PLS regression for NIR spectroscopy in Python
Not every data point is created equal. In this post we'll show how to perform outliers detection with PLS regression for NIR spectroscopy in Python.
Exporting NIR regression models built in Python
Data Operations and Plotting
08/19/2018
Three methods to export a Python NIR regression model and how to load it back for future use. Worked Python codes to discuss pros and cons of these methods.
Two scatter correction techniques for NIR spectroscopy in Python
Worked example of two scatter correction techniques for NIR spectroscopy in Python: Multiplicative Scatter Correction and Standard Normal Variate.
A variable selection method for PLS in Python
Improve the quality of your PLS regression using variable selection. This tutorial will work through a variable selection method for PLS in Python.
Partial Least Squares Regression in Python
Partial Least Squares Regression, Regression
06/14/2018
Step by step tutorial on how to build a NIR calibration model using Partial Least Squares Regression in Python.
Principal Component Regression in Python
Principal Components Regression, Regression
05/12/2018
An in-depth introduction to Principal Component Regression in Python using NIR data. PCR is the combination of PCA with linear regression. Check it out.
Classification of NIR spectra using Principal Component Analysis in Python
Classification, Principal Components Analysis
03/23/2018
An in-depth tutorial on how to run a classification of NIR spectra using Principal Component Analysis in Python. Step by step example with code.
NIR classification of macadamia kernels
Classification, Principal Components Analysis, Use Cases
07/06/2017
Can we use NIR analysis to grade macadamias? Check out our preliminary results of NIR classification of macadamia kernels using Principal Component Analysis.
The cricket thermometer: an introduction to Principal Component Analysis in Python
Classification, Principal Components Analysis
03/21/2017
A worked example for an introduction to Principal Component Analysis in Python.
