itineRis

R package developed for statistical analysis isotops developped in the frame of the ANR-Itineris

Install and load package

Install the R package

devtools::install_github("zoometh/itineRis")

And load it

library(itineRis)

By default, output will be saved in the results/ folder. You can change the output folder by changing the dirOut option in the various functions.

Correspondence Analysis

Run the isotop_ca() function to compute Correspondence Analysis (CA)

isotop_ca(col.group = "Provenienza",
          pt_siz = 2,
          lbl.size = 3)

The user can select 3 or more variables that appears both in the dataset and in the reference thesaurus. The latter gathers lists of elements (Cu, Ar, …), chemical compounds (SiO2, TiO2, …), or isotopes.

Selection of variables (elements)

The result will is:

CA on elements

interactive ternary plot of LIA

Create an interactive ternary plot, with plotly, to visualise the results of lead isotopic analysis (LIA)

Create an hash object to store

d <- hash::hash()

Read the LIA measures for objects and mines

d <- read_lia(d = d,
              d.tag = "lia.objects",
              df.path = "C:/Rprojects/itineRis/results/LIA data objects.xlsx",
              header.line = 2)
d <- read_lia(d = d,
              d.tag = "lia.mines",
              pattern.objects.num = "Locality/.Mine",
              pattern.objects.Pb206_Pb204 = "206Pb/204Pb",
              pattern.objects.Pb207_Pb204 = "207Pb/204Pb",
              pattern.objects.Pb208_Pb204 = "208Pb/204Pb",
              df.path = "C:/Rprojects/itineRis/results/Coordinates-mines_Thomas Huet.xlsx",
              header.line = 1)

Group the datasets, assign colors and symbols

library(dplyr)

mydf <- rbind(d$lia.mines, d$lia.objects)
df.isotop <- isotop_dataframe(df = mydf, df.path = NA)
df.isotop[df.isotop$object == "golasecca", "color.object"] <- "#0000FF"
df.isotop[df.isotop$object == "hochdorf", "color.object"] <- "#00FFFF"
df.isotop[df.isotop$object == "France", "color.object"] <- "#FF0000"
df.isotop[df.isotop$object == "France", "symbol"] <- "triangle"
df.isotop[df.isotop$object == "Iberian Peninsula", "color.object"] <- "#FFA500"
df.isotop[df.isotop$object == "Iberian Peninsula", "symbol"] <- "triangle"
df.isotop[df.isotop$object == "Switzerland", "color.object"] <- "#964B00"
df.isotop[df.isotop$object == "Switzerland", "symbol"] <- "triangle"

Create a ternary plot

isotop_ternaryplot(df.isotop)

landing page of the interactive Plotly plot

In Shiny

A similar function is hosted on a Shiny Server: http://shinyserver.cfs.unipi.it:3838/teach/stats/stats/dim3/

img-name
landing page of the interactive Shiny ternary plot

The parameters (objects, mines, kmeans) can be selected with the input box:

objects: categories of objects (by default golasecca or hordorf)
mines: mines’ countries (by default France, Iberian Peninsula, or Switzerland)
kmeans: number of Kmeans¹ centers, or clusters (by default between 1 to 5)

img-name
input selection

Use the Zoom tool to select a region of interest in the graph (to zoom out, double click).

img-name
Zoom in

The Zoom tool, the Span tool and the Hover info allow you to view the results sample by sample.

img-name
hover info for the sample `4230_08` allocated to the cluster `1`

Each sample is tagged with the number of the Kmeans cluster to which it belongs.

Interactive map of mines

Create an interactive leaflet map of mines

d <- hash::hash()
d <- read_mines(d = d,
                df.path = "C:/Rprojects/itineRis/results/Coordinates-mines_Thomas Huet.xlsx")
map_leaflet(d = d,
            d.coords = "mines.coords",
            export.plot = T,
            out.plot = "map_sites.html",
            dirOut = "C:/Rdev-itineris/itineris/data/")

Gives:

https://zoometh.github.io/itineRis/results/map_sites.html

Thesaurus

The function thesaurus() creates interactive graphs to display the thesaurus of the project

thes <- "https://raw.githubusercontent.com/ANR-Itineris/itineris/main/lod/thesaurus/Itineris_th101.rdf"
thesaurus(root = "artefact",
          outPlot = "collapsibleTree",
          export.plot = T,
          outDir = "C:/Rprojects/itineRis/results/")

https://zoometh.github.io/itineRis/results/artefact

the Kmeans (R function kmeans()) is run on the selected objects and mines. It tags every sample with the number of its cluster () ↩

This site is open source. Improve this page.