Soil physical, chemical and fertility properties from 50 samples in the Ucayali region of Peru. This is the package's general-purpose example dataset, used throughout the documentation and vignettes.
Format
A data frame with 50 rows and 16 columns:
- SampleID
Character. Unique identifier for each soil sample (UCY001-UCY050)
- Sand
Numeric. Sand content as percentage of soil texture (0-100)
- Silt
Numeric. Silt content as percentage of soil texture (0-100)
- Clay
Numeric. Clay content as percentage of soil texture (0-100)
- BD
Numeric. Bulk density in g/cm³
- pH
Numeric. Soil pH measured in water (1:1)
- OM
Numeric. Organic matter content as percentage
- SOC
Numeric. Soil organic carbon content as percentage
- N
Numeric. Total nitrogen content as percentage
- P
Numeric. Available phosphorus in mg/kg (Olsen method)
- K
Numeric. Exchangeable potassium in mg/kg
- CEC
Numeric. Cation exchange capacity in cmol/kg
- Ca
Numeric. Exchangeable calcium in cmol/kg
- Mg
Numeric. Exchangeable magnesium in cmol/kg
- EC
Numeric. Electrical conductivity in dS/m
- S
Numeric. Sulfur content in mg/kg
Details
The soils are generally acidic (pH 4-7.5), with moderate organic matter and variable nutrient levels typical of tropical agricultural soils in the Ucayali region.
Choosing between the two example datasets. The values here are simulated with each property drawn independently, so the dataset has almost no covariance structure: the largest off-diagonal Spearman correlation is 0.66 and exactly one pair of indicators clears a 0.6 threshold. That is fine, and realistic enough, for demonstrating scoring, weighting and aggregation, which treat indicators one at a time.
It is not suitable for any method that reads the structure
between indicators – correlation-network selection, functional
grouping, redundancy screening. Use soil_structured for those.
A note on this dataset's own soil quality index
Running sqi_validate on the default recipe applied to this
data fires the middle-band warning: the resulting index spans roughly
0.36 to 0.70, so every sample lands in the middle bands and both extreme
categories are empty. That is the diagnostic working, not a defect, and it
is a useful thing to see before trusting an index built the same way on
real data.
See also
soil_structured, which carries realistic covariance
and is the one to use for indicator-selection methods
Examples
# Load the dataset
data(soil_data)
# View structure
str(soil_data)
#> 'data.frame': 50 obs. of 16 variables:
#> $ SampleID: chr "UCY001" "UCY002" "UCY003" "UCY004" ...
#> $ Sand : num 42.4 42.7 54.6 41.9 50.1 49.5 57.6 38.3 40.8 38.6 ...
#> $ Silt : num 35.4 30.2 25.4 37.4 30.3 31.8 20.1 44.6 34.4 30.8 ...
#> $ Clay : num 22.2 27.1 20 20.7 19.6 18.7 22.3 17.1 24.8 30.6 ...
#> $ BD : num 1.47 1.47 1.4 1.2 1.33 1.31 1.43 1.29 1.5 1.29 ...
#> $ pH : num 7 6.2 5 5.6 4.9 4.8 4.6 4.7 6.5 5.2 ...
#> $ OM : num 2.46 2.29 2.49 2.88 4.24 2.72 3.77 3.37 2.7 1.42 ...
#> $ SOC : num 1.24 1.22 1.13 1.07 1.38 1.77 0.59 1.71 2.22 2.62 ...
#> $ N : num 0.201 0.05 0.129 0.156 0.105 0.167 0.171 0.148 0.05 0.279 ...
#> $ P : num 7.8 4.5 6.1 7.9 10 3 7 10.3 6.4 8.7 ...
#> $ K : int 131 121 106 113 118 40 123 83 101 88 ...
#> $ CEC : num 10.2 9 15.1 14.3 7.5 11.7 9.3 5.8 12.5 11.8 ...
#> $ Ca : num 5.51 3.56 5.83 4.8 6.48 2.75 4.44 3.87 4.26 3.74 ...
#> $ Mg : num 2.44 1.78 1.78 0.89 2.27 1.67 1.41 0.95 1.62 1.29 ...
#> $ EC : num 0.06 0.26 0.31 0.1 0.3 0.17 0.35 0.3 0.33 0.26 ...
#> $ S : num 5.54 7.2 8.5 6.5 15.5 ...
# Summary statistics
summary(soil_data)
#> SampleID Sand Silt Clay BD
#> Length :50 Min. :29.60 Min. :15.90 Min. :11.30 Min. :1.150
#> N.unique :50 1st Qu.:39.25 1st Qu.:25.52 1st Qu.:18.90 1st Qu.:1.280
#> N.blank : 0 Median :43.45 Median :30.65 Median :23.35 Median :1.335
#> Min.nchar: 6 Mean :45.21 Mean :31.45 Mean :23.34 Mean :1.355
#> Max.nchar: 6 3rd Qu.:50.55 3rd Qu.:36.17 3rd Qu.:27.68 3rd Qu.:1.430
#> Max. :64.30 Max. :46.30 Max. :38.00 Max. :1.800
#> pH OM SOC N
#> Min. :4.100 Min. :1.220 Min. :0.570 Min. :0.0500
#> 1st Qu.:4.600 1st Qu.:2.507 1st Qu.:1.240 1st Qu.:0.1128
#> Median :5.050 Median :2.895 Median :1.585 Median :0.1505
#> Mean :5.192 Mean :3.024 Mean :1.601 Mean :0.1469
#> 3rd Qu.:5.600 3rd Qu.:3.643 3rd Qu.:1.972 3rd Qu.:0.1755
#> Max. :7.000 Max. :4.860 Max. :2.620 Max. :0.2790
#> P K CEC Ca
#> Min. : 2.000 Min. : 40.0 Min. : 5.80 Min. :1.130
#> 1st Qu.: 6.150 1st Qu.: 88.0 1st Qu.:10.22 1st Qu.:3.560
#> Median : 8.050 Median :102.0 Median :11.75 Median :4.410
#> Mean : 7.860 Mean :102.2 Mean :12.08 Mean :4.366
#> 3rd Qu.: 9.575 3rd Qu.:118.0 3rd Qu.:14.05 3rd Qu.:5.178
#> Max. :15.200 Max. :144.0 Max. :19.30 Max. :6.730
#> Mg EC S
#> Min. :0.240 Min. :0.0500 Min. : 5.540
#> 1st Qu.:1.292 1st Qu.:0.1700 1st Qu.: 8.852
#> Median :1.700 Median :0.2400 Median : 9.640
#> Mean :1.701 Mean :0.2374 Mean :10.138
#> 3rd Qu.:2.180 3rd Qu.:0.3000 3rd Qu.:10.807
#> Max. :3.050 Max. :0.5000 Max. :18.560
# Basic SQI calculation with extended properties
if (FALSE) { # \dontrun{
result <- compute_sqi_properties(
data = soil_data,
properties = c("pH", "OM", "N", "P", "K", "S", "CEC"),
id_column = "SampleID"
)
print(result)
} # }