Skip to content

HigherPasserines

Cone Fundamentals for the Higher Passerines

Bundled with colorSpec · exported, loads via data()

Class
colorSpec
colorSpec/matrix
Dimensions
401 × 4
Missing
0
Format
rda · xz
version 3 · 108 KB
File
data/colorSpec.rda
Fingerprint
6da76a28
Format documented
Yes
Source
`https://onlinelibrary.wiley.com/doi/full/10.1111/j.1095-8312.2005.00540.x`
About this object
A flat block of values with dimensionsA block of cells sharing every edge, with nothing above the columns and nothing in it: a name here is hung on the whole block rather than on one column.

A single vector with a dim attribute. All elements share one type. R object catalogue

Values

Computed over every cell in the object, reading it as one vector.

numeric
Cells
1,604
Missing
0
Min.
0
1st Qu.
0
Median
8.0e-8
Mean
0.00249377
3rd Qu.
0.00172026
Max.
0.0194288
Std. dev
0.00487338
Distinct
784
Skewness
1.909
Kurtosis
2.308
Zeros
728 45.4%
Outliers
321 0 low · 321 high
Margins

The mean of each row, and of each column, summarised. Reading every cell as one vector gives a matrix and its transpose the same answer; these do not.

MarginMin.1st Qu.MedianMean3rd Qu.Max.Std. dev
Rows5.9725e-50.00129180.00275080.00249380.00350980.00489580.0013422

The row means vary far more than the column means, so what separates the values runs down the rows.

Dimensions and labels
MarginLabelsCount
ColumnsUV, Short, Medium, Long4
Declared attributes
Other attributes

Read by name, kind and length, with the value where it is short enough to show. No rule here knows what any of them means.

AttributeHoldsLengthValue
metadatalist3
quantitycharacter1photons->neural
step.wlnumeric11
wavelengthnumeric401300 – 700
Columns

This object is 4 columns wide. A matrix carries no per-column schema, so its width is what the catalogue records and there is nothing to list one column at a time. What it holds is described whole instead, summarised over every cell: see Values and Margins above.

Structure fingerprint

333 other datasets carry the same structure fingerprint. That fingerprint is a digest of the column names and their types, and a matrix carries neither, so on one it comes down to the width and the type held: a match is another object of 4 columns holding the same type rather than one with the same columns. The datasets are not listed, because that is not a list of anything this one is related to.

Version history
VersionReleasedSizeChangeStored asConfidence
1.8-02025-06-10401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.7-02025-02-12401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.6-02025-01-15401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.5-02024-01-28401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.4-02022-05-04401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.3-02021-12-20401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.2-12020-04-01401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.1-12019-12-07401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
1.0-12019-06-24401 × 4Unchangedrda · xz
version 3 · 108 KB
exact
0.9-12019-05-31401 × 4Repackagedrda · xz
version 3 · 108 KB
exact
0.8-32019-03-04401 × 4Unchangedrda · xz
version 2 · 107 KB
exact
0.7-52018-11-19401 × 4Unchangedrda · xz
version 2 · 99 KB
exact
0.7-32018-04-02401 × 4Unchangedrda · xz
version 2 · 99 KB
exact
0.6-22017-12-04401 × 4Unchangedrda · xz
version 2 · 92 KB
exact
0.6-12017-11-17401 × 4Unchangedrda · xz
version 2 · 92 KB
exact
0.5-32016-05-17401 × 4Unchangedrda · xz
version 2 · 92 KB
exact
0.5-22016-05-15401 × 4–rda · xz
version 2 · 92 KB
exact

Privacy choices

These apply to this browser and are stored on this device only. Nothing about your choice is sent to us.

Read the privacy policy