update vignette

[priviere]

• ecovalence add that is on gxe + résiduals

• in appendices : example with apply and a vector of variables for GxE for example

• add decision tree from Isabelle

• differenciate memory needed for the different analyses

[priviere]

done in v0.18 :

• ecovalence add that is on gxe + résiduals
• add decision tree

still to do :

• in appendices : example with apply and a vector of variables for GxE for example
• differenciate memory needed for the different analyses

[priviere]

done in v0.22 :

• example with lapply for GxE, spatial, model 1, model 2
• integrate multivariate.R in workflow fig + table
• describe data_version in format section
• for each section, add a paragraph format data

still to do :

• explain Skewness and Kurtosis test for GxE and Spatial
• add section on critics regarding big data and data mining in the context of PPB programme
• write frequentist statistics theory part
• write sow measure harvest section
• differentiate memory needed for the different analyses, see https://github.com/tpq/peakRAM
• translate the vignette to Rmarkdow in order to add md files into shiny app (#64) and be more flexible in the output
• improve translation procedure for contribution vignette http://developer.r-project.org/Translations.html https://stackoverflow.com/questions/29093673/how-to-translate-package-content See library(tools); update_pkg_po("./")

[priviere]

done in v0.23 :

• translate the vignette to Rmarkdow (and bookdown) in order to add md files into shiny app (#64) and be more flexible in the output
• explain Skewness and Kurtosis test for GxE and Spatial

still to do :

• add in contribution how to run bookdown, etc (cf dashboard file on my computer)
• cf here to add info on tests
• add section on critics regarding big data and data mining in the context of PPB programme
• complete frequentist and bayesian statistics theory part (add example with graphs ?)
• write sow measure harvest section: cf mail sent to Sophie et Gaelle 04 01 2018 17:10
• differentiate memory needed for the different analyses, see https://github.com/tpq/peakRAM
• improve translation procedure for contribution vignette http://developer.r-project.org/Translations.html https://stackoverflow.com/questions/29093673/how-to-translate-package-content See library(tools); update_pkg_po("./")
• compare analysis, e.g. for each analysis, see the df of the error for example
• add question to the objective
• mention help regarding plot with ?plot.PPBstats
• cf Facundo comments : There is some confusion of frequentist (statistical significance) and Bayesian (posterior probabilities) concepts here. Statistical significance refers to the variability of some estimator (e.g. $\mu{ij} - \mu{i'j}$) under replications of the same experiment (i.e. different datasets), while the posterior probability quantifies the information about some variable given the current dataset. You can certainly establish some thershold $\alpha$ so that whenever the posterior probability of $H_0$ or $H_1$ exceeds $1 - \alpha$ you can claim that the data provides evidence of some difference (not necessarily large) between groups. But calling that significant is not correct.
• change GGE formula: cf mail à Isa le 11 05 2018 14:25, change code in consequence
• agro intro: add argument of correlated response to selection (cf PhD Pierre)
• cf Be-Breeder: an R/Shiny application for phenotypic data analyses in plant breeding http://www.scielo.br/scielo.php?pid=S1984-70332018000200241&script=sci_arttext
• create a logo PPBstats ?!?

[priviere]

done in v0.25 :

• write sow measure harvest section: cf mail sent to Sophie et Gaelle 04 01 2018 17:10
• agro worflow: make a general one in the introduction and make one workflow per family of analysis in order to make it more readable and flexible
• add in contribution how to run bookdown, etc (cf dashboard file on my computer)
• add section on critics regarding big data and data mining in the context of PPB programme in "communicate to farmers" section
• cf Facundo comments : There is some confusion of frequentist (statistical significance) and Bayesian (posterior probabilities) concepts here. Statistical significance refers to the variability of some estimator (e.g. $\mu{ij} - \mu{i'j}$) under replications of the same experiment (i.e. different datasets), while the posterior probability quantifies the information about some variable given the current dataset. You can certainly establish some thershold $\alpha$ so that whenever the posterior probability of $H_0$ or $H_1$ exceeds $1 - \alpha$ you can claim that the data provides evidence of some difference (not necessarily large) between groups. But calling that significant is not correct.
• mention help : Note that check_model(), mean_comparison(), biplot_data() and plot() are S3 method. Therefore, you should type ?check_model, ?mean_comparison, ?biplot_data or ?plot.PPBstats to have general features and then see in details for specific functions.

still to do:

• cf here to add info on tests
• complete frequentist and bayesian statistics theory part (add example with graphs ?)
• change GGE formula: cf mail à Isa le 11 05 2018 14:25, change code in consequence
• agro intro: add argument of correlated response to selection (cf PhD Pierre)
• create a logo PPBstats
• add a section publications on the website with articles or reports that cites PPBstats
  • http://www.diversifood.eu/wp-content/uploads/2018/11/Diversifood_11_smart_methods.pdf
  • http://www.diversifood.eu/wp-content/uploads/2018/12/booklet3_decentralized_on_farm_breeding_BAT_web_A4_2.pdf
  • section 5 http://www.diversifood.eu/wp-content/uploads/2019/02/booklet6_innovative_approaches_BAT_web_v2.pdf
  • https://link.springer.com/article/10.1007%2Fs10681-019-2447-9

[priviere]

done in v0.26 :

• add a section publications on the website with articles or reports that cite PPBstats

still to do:

• cf here to add info on tests
• complete frequentist and bayesian statistics theory part (add example with graphs ?)
• change GGE formula: cf mail à Isa le 11 05 2018 14:25, change code in consequence
• agro intro: add argument of correlated response to selection (cf PhD Pierre)
• create a logo PPBstats
• typo: a R package, and not an R package ...
• add info regarding sowing practices: cf Rebetzke et al. (2014) provide a review on plot size, focusing mainly on wheat in Australia. Differences in height between cultivars caused neighbor effects. Plots with four rows of wheat showed a 10% bias in one example, reversing the ranking of varieties. They discuss how under drought more competitive varieties outyielded more drought-tolerant varieties. They recommend plots with at least 6 rows for accurate yield assessment in the case of wheat. Omitting border rows can further help to reduce biases arising from border effects. Border effects due to plant height are less accentuated at lower latitudes as shadowing is less important. Also, competition is less important in areas where resources for water or nutrients are not the main yield limitation, but heat or cold stress, or pests and diseases. Omitting border rows is often a standard practice in on-farm trials. Rebetzke, Greg J., Ralph Tony A. Fischer, Anthony F. van Herwaarden, Dave G. Bonnett, Karine Chenu, Allan R. Rattey, and Neil A. Fettell. 2014. “Plot Size Matters: Interference from Intergenotypic Competition in Plant Phenotyping Studies.” Functional Plant Biology: FPB 41 (2): 107–18. https://doi.org/10.1071/FP13177.