Skip to contents

Classify a numeric column into map classes

geo_breaks.Rd

geo_breaks() returns class breaks for a numeric vector, and geo_classify() adds the class of every row to a data frame, ready for a choropleth or a bubble map. The styles follow tmap and classInt:

  • "quantile": equal numbers of observations per class;

  • "equal": equal-width classes over the range;

  • "jenks": Fisher-Jenks natural breaks, which minimise the within-class sum of squares exactly (Fisher 1958; computed in C++). The breaks equal classInt's "jenks" style; classInt's "fisher" style gives the same classes with breaks halfway between neighbouring values;

  • "pretty": rounded breaks from base::pretty();

  • "sd": breaks at the mean plus or minus whole standard deviations;

  • "fixed": the breaks you give.

Like GeoSpaX's AutoMap, geo_classify() explains its choice: the result carries the breaks, the number of rows in each class and the goodness of variance fit (GVF, 1 = classes explain all variance) in attribute "classification".

Usage

geo_breaks(
  x,
  n = 5,
  style = c("quantile", "equal", "jenks", "pretty", "sd", "fixed"),
  breaks = NULL
)

geo_classify(
  data,
  value,
  n = 5,
  style = c("quantile", "equal", "jenks", "pretty", "sd", "fixed"),
  breaks = NULL,
  digits = 3,
  name = NULL
)

Arguments

x

A numeric vector.

n

Number of classes.

style

Classification style (see Description).

breaks

For style = "fixed", the full vector of breaks, including the minimum and maximum.

data

A data frame or sf table.

value

<data-masking> Numeric column to classify.

digits

Significant digits used in class labels.

name

Name of the new class column (default "<value>_class").

Value

geo_breaks() returns the sorted breaks (n + 1 values, minimum to maximum). geo_classify() returns data with an ordered factor column of class labels such as "10 - 20".

References

Fisher, W. D. (1958). On grouping for maximum homogeneity. Journal of the American Statistical Association, 53, 789-798. doi:10.1080/01621459.1958.10501479

Jenks, G. F. (1977). Optimal data classification for choropleth maps. Occasional paper 2, Department of Geography, University of Kansas.

Examples

set.seed(1)
x <- c(rnorm(20, 10), rnorm(20, 30), rnorm(20, 60))
geo_breaks(x, n = 3, style = "jenks")
#> [1]  7.78530 11.59528 31.35868 61.98040

sites <- geo_example_sites()
geo_classify(sites, elevation, n = 3, style = "equal")
#> # A tibble: 8 × 7
#>   site     lon   lat elevation elevation_band coordinate_note    elevation_class
#>   <chr>  <dbl> <dbl>     <int> <fct>          <chr>              <ord>          
#> 1 MW0200  145. -5.74       200 200 m          published bait GP… 200 - 1370     
#> 2 MW0700  145. -5.73       700 700 m          published bait GP… 200 - 1370     
#> 3 MW1200  145. -5.72      1200 1200 m         published bait GP… 200 - 1370     
#> 4 MW1700  145. -5.76      1700 1700 m         published bait GP… 1370 - 2530    
#> 5 MW2200  145. -5.76      2200 2200 m         published bait GP… 1370 - 2530    
#> 6 MW2700  145. -5.81      2700 2700 m         published bait GP… 2530 - 3700    
#> 7 MW3200  145. -5.80      3200 3200 m         published bait GP… 2530 - 3700    
#> 8 MW3700  145. -5.78      3700 3700 m         published bait GP… 2530 - 3700