Learn R Programming

munsellinterpol (version 3.2-0)

roundHVC: Round a Given Munsell HVC to the Closest Sample in One or More Books of Samples

Description

The functions in this package compute Munsell HVC coordinates with high precision. When reporting the Munsell notation it is often desirable, for historical consistency or other reasons, to report a notation that is actually appears in one or more published books of Munsell samples. This rounding operation requires exact knowledge of such books. We use a paper by Ferguson, which gives the contents of books on these fields: soil, rocks, beads, and plants. It also includes The New Munsell Student Color Set, see References.

Usage

roundHVC( HVC, books )

Value

A data.frame with a N rows and these columns:

HVC

the original Nx3 input matrix HVC.

ISCC-NBS Name

the ISCC-NBS Name (the color name) of the given input sample.

MunsellRounded

the rounded Munsell notation that is rounded from HVC. It is guaranteed to be in one of the books given by the argument books.

Ferguson Name

the color name of MunsellRounded, as assigned by Ferguson

The row names are set to the row names of HVC. But if these are NULL they are set to the Munsell notation of the input HVC. Finally, if these row names have duplicates, they are set to 1:N.

Arguments

HVC

a numeric Nx3 matrix with Munsell HVC values in the rows, or a vector that can be converted to such a matrix, by row.
HVC can also be a character N-vector with Munsell Notations, which is converted to an Nx3 matrix using HVCfromMunsellName().

books

a comma-delimted string that specifies one or more books of samples. The words in the string can be 'soil', 'rock', 'bead', 'plant', or 'newstudent'. Matching is partial (initial substring only) and case-insensitive. See References and Examples.

Acknowledgements

The conversion of the book data to Supplement1_3.6.2024.csv was done by Willie Ondricek, who provided all the motivation for this function. He also kindly provided photographs of the soil book, and much other material.

Author

Glenn Davis

Details

First, the set of samples over all selected books is formed. Then the closest sample in this set to the given notation is found. For the color distance, the function NickersonColorDifference() is used. Unfortunately, the particular coefficients in this difference equation can lead to large regions of Munsell color space with exact ties for the two closest samples. These exact ties make the final result unpredictable and sometime unintuitive. As a tie-breaker, a very small multiple (1.e-6) of plain Eucliden distance is added to the Nickerson difference, which has proven effective.

References

Ferguson, Jonathan. Munsell notations and color names: Recommendations for archaeological practice. Journal of Field Archaeology. 39:4. 327-335. (2014). Online Supplement #1.

Long, J. T. The New Munsell Student Color Set. New York: Fairchild Books. 2011.

Munsell Color. Munsell Soil-Color Charts. Grand Rapids, MI: 2009.

Munsell Color. Geological Rock-Color Chart. GrandRapids, MI: 2009.

Munsell Color. Munsell Bead Color Book. Grand Rapids, MI: 2012.

Munsell Color. Munsell Color Charts for Plant Tissues. New Windsor, NY: 1977.

Examples

Run this code
## search the soil book for the best match for 3 samples
roundHVC( c("7.7YR 3.4/6.1", "2.6PB 6.1/4.5", "N 6.6/" ), books='soil' )

##               HVC.H HVC.V HVC.C ISCC-NBS Name MunsellRounded Ferguson Name
## 7.7YR 3.4/6.1  17.7   3.4   6.1  strong brown      7.5YR 4/6  Strong brown
## 2.6PB 6.1/4.5  72.6   6.1   4.5     pale blue        5PB 6/1   Bluish gray
## N 6.6/          0.0   6.6   0.0    light gray           N 7/    Light gray

## The middle sample has a poor match, since "purplish-blue" with Chroma=4.5
## is not well-matched by soil color samples.
## Search again, but this time in all 5 books.

roundHVC( c("7.7YR 3.4/6.1", "2.6PB 6.1/4.5", "N 6.6/" ), books='s,b,r,p,n' )

##               HVC.H HVC.V HVC.C ISCC-NBS Name MunsellRounded Ferguson Name
## 7.7YR 3.4/6.1  17.7   3.4   6.1  strong brown      7.5YR 4/6  Strong brown
## 2.6PB 6.1/4.5  72.6   6.1   4.5     pale blue        5PB 6/4     Pale blue
## N 6.6/          0.0   6.6   0.0    light gray           N 7/    Light gray

## This time, for the middle sample, a better match was found.
## The other 2 sample are unchanged.

Run the code above in your browser using DataLab