Pairwise comparisons in nonlinear regression

Published at January 19, 2021 ·  6 min read

Pairwise comparisons are one of the most debated topic in agricultural research: they are very often used and, sometimes, abused, in literature. I have nothing against the appropriate use of this very useful technique and, for those who are interested, some colleagues and I have given a bunch of (hopefully) useful suggestions in a paper, a few years ago (follow this link here).

Pairwise comparisons usually follow the application of some sort of linear or generalised linear model; in this setting, the ‘emmeans’ package (Lenth, 2020) is very handy, as it uses a very logical approach. However, we can find ourselves in the need of making pairwise comparisons between the elements of a vector, which does not came as the result of linear model fitting.

lmDiallel: a new R package to fit diallel models. The Griffing's models (1956)

Published at January 12, 2021 ·  10 min read

Diallel mating designs are often used by plant breeders to compare the possible crosses between a set of genotypes. In spite of such widespread usage, the process of data analysis in R is not yet strightforward and it is not clear which tool should be routinely used. We recently gave a small contribution by publishing a paper in Plant Breeding (Onofri et al., 2020 ), where we advocated the idea that models for diallel crosses are just a class of general linear models, that should be fit by Ordinary Least Squares (OLS) or REstricted Maximum Likelihood methods (REML).

lmDiallel: a new R package to fit diallel models. The Hayman's model (type 2)

Published at January 5, 2021 ·  9 min read

This posts follows two other previously published posts, where we presented our new ‘lmDiallel’ package (see here) and showed how we can use it to fit the Hayman’s model type 1, as proposed in Hayman (1954) (see here). In this post, we will give a further example relating to another very widespread model from the same author, the Hayman’s model type 2. We apologise for some overlapping with previous posts: we think this is necessary so that each post can be read on its own.

General code to fit ANOVA models with JAGS and 'rjags'

Published at December 23, 2020 ·  15 min read

One of the reasons why I like BUGS and all related dialects has been put nicely in a very good book, i.e. “Introduction to WinBUGS for ecologists” (Kery, 2010); at page 11, the author says: “WinBUGS helps free the modeler in you”. Ultimately, that statement is true: when I have fully understood a model with all its components (and thus I have become a modeler), I can very logically translate it to BUGS code. The drawback is that, very often, the final coding appears to be rather ‘problem-specific’ and difficult to be reused in other situations, without an extensive editing work.

From ''for()'' loops to the ''split-apply-combine'' paradigm for column-wise tasks: the transition for a dinosaur

Published at December 11, 2020 ·  9 min read

I have been involved with data crunching for 30 years, and, due to my age, I see myself as a dinosaur within the R-users community. I must admit, I’m rather slow to incorporate new paradigms in my programming workflow … I’m pretty busy and the time I save today is often more important than the time I could save in the future, by picking up new techniques. However, resisting to progress is not necessarily a good idea and, from time to time, also a dinosaur feels like living more dangerously and exploring new ideas and views.

Accounting for the experimental design in linear/nonlinear regression analyses

Published at December 4, 2020 ·  11 min read

In this post, I am going to talk about an issue that is often overlooked by agronomists and biologists. The point is that field experiments are very often laid down in blocks, using split-plot designs, strip-plot designs or other types of designs with grouping factors (blocks, main-plots, sub-plots). We know that these grouping factors should be appropriately accounted for in data analyses: ‘analyze them as you have randomized them’ is a common saying attributed to Ronald Fisher. Indeed, observations in the same group are correlated, as they are more alike than observations in different groups. What happens if we neglect the grouping factors? We break the independence assumption and our inferences are invalid (Onofri et al., 2010).

lmDiallel: a new R package to fit diallel models. The Hayman's model (type 1)

Published at November 26, 2020 ·  15 min read

In a previous post we have presented our new ‘lmDiallel’ package (see this link here and see also the original paper in Theoretical and Applied Genetics). This package provides an extensions to fit a class of linear models of interest for plant breeders or geneticists, the so-called diallel models. In this post and other future posts we would like to present some examples of how to use this package: please, sit back and relax and, if you have comments, let us know, using the email link at the bottom of this post.

lmDiallel: a new R package to fit diallel models. Introduction

Published at November 11, 2020 ·  7 min read

Together with some colleagues from the plant breeding group, we have just published a new paper, where we presented a bunch of R functions to analyse the data from diallel experiments. The paper is titled ‘Linear models for diallel crosses: a review with R functions’ and it is published in the ‘Theoretical and Applied Genetics’ Journal. If you are interested, you can take a look here at this link.

Diallel experiments are based on a set of possible crosses between some homozygous (inbred) lines. For example, if we have the male lines A, B and C and the female lines A, B and C (same lines used, alternatively, as male and female), we would have the following selfed parents: AA, BB and CC and the following crosses: AB, AC, BC. In some instances, we might also have the reciprocals BA, CA and CB. Selfed parents and crosses are compared on a Randomised Complete Block Design, usually replicated across seasons and/or locations.

QQ-plots and Box-Whisker plots: where do they come from?

Published at October 15, 2020 ·  7 min read

Building ANOVA-models for long-term experiments in agriculture

Published at August 20, 2020 ·  29 min read

This is the follow-up of a manuscript that we (some colleagues and I) have published in 2016 in the European Journal of Agronomy (Onofri et al., 2016). I thought that it might be a good idea to rework some concepts to make them less formal, simpler to follow and more closely related to the implementation with R. Please, be patient: this lesson may be longer than usual.

First question: what are long-term experiments? Agricultural experiments have to deal with long-term effects of cropping practices. Think about fertilisation: certain types of organic fertilisers may give effects on soil fertility, which are only observed after a relatively high number of years (say: 10-15). In order to observe those long-term effects, we need to plan Long Term Experiments (LTEs), wherein each plot is regarded as a small cropping system, with the selected combination of rotation, fertilisation, weed control and other cropping practices. Due to the fact that yield and other relevant variables are repeatedly recorded over time, LTEs represent a particular class of multi-environment experiments with repeated measures.