Pottery2: Chemical Analysis of Romano-British Pottery

Description Usage Format Details Source References See Also Examples

Description

Results of chemical analyses of 48 specimens of Romano-British pottery published by Tubb et al. (1980). The numbers are the percentage of various metal oxides found in each sample for elements of concentrations greater than 0.01%. This is the original data set from Tubb et al. (1980), in contrast to Pottery.

Usage

1

Format

A data frame with 48 observations on the following 12 variables.

Region

a factor with levels Gl NF Wales

Site

a factor with levels AshleyRails Caldicot Gloucester IsleThorns Llanedryn

Kiln

a factor with levels 1 2 3 4 5

Al

amount of aluminum oxide, Al_2O_3

Fe

amount of iron oxide, Fe_2O_3

Mg

amount of magnesium oxide, MgO

Ca

amount of calcium oxide, CaO

Na

amount of sodium oxide, Na_2O

K

amount of potassium oxide, K_2O

Ti

amount of titanium oxide, TiO_2

Mn

amount of manganese oxide, MnO

Ba

amount of BaO

Details

The specimens are identified by their rownames in the data frame. Kiln indicates at which kiln site the pottery was found; Site gives the location names of those sites. The kiln sites come from three Regions, ("Gl"=1, "Wales"=(2, 3), "NF"=(4, 5)), where the full names are "Gloucester", "Wales", and "New Forrest".

The variable Kiln comes pre-supplied with contrasts to test interesting hypotheses related to Site and Region.

Source

Slightly modified from files found at: http://people.tamu.edu/~dcarlson/quant/data/RBPottery.html

References

Baxter, M. J. 2003. Statistics in Archaeology. Arnold, London.

Tubb, A., A. J. Parker, and G. Nickless. 1980. The Analysis of Romano-British Pottery by Atomic Absorption Spectrophotometry. Archaeometry, 22, 153-171.

See Also

Pottery for the related (subset) data set

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
data(Pottery2)
# contrasts for Kiln correspond to between Region [,1:2] and within Region [,3:4]
contrasts(Pottery2$Kiln)

pmod <-lm(cbind(Al,Fe,Mg,Ca,Na,K,Ti,Mn,Ba)~Kiln, data=Pottery2)
Anova(pmod)

# extract coefficient names for linearHypotheses
coefs <- rownames(coef(pmod))[-1]

# test differences among regions
linearHypothesis(pmod, coefs[1:2])
# test differences within regions B, C
linearHypothesis(pmod, coefs[3:4])

heplot(pmod, fill=c(TRUE,FALSE), hypotheses=list("Region" =coefs[1:2], "WithinBC"=coefs[3:4]))

# all pairwise views;  note that Ba shows no effect
pairs(pmod, fill=c(TRUE,FALSE))

# canonical view, via candisc::heplot

if (require(candisc)) {

# canonical analysis: how many dimensions?
(pcan <- candisc(pmod))

heplot(pcan, scale=18, fill=c(TRUE,FALSE), var.col="darkgreen", var.lwd=2, var.cex=1.5)

## Not run: 
heplot3d(pcan, scale=8)

## End(Not run)
}

heplots documentation built on Oct. 7, 2021, 1:07 a.m.