Type:	Package
Title:	Comprehensive Soil Quality Index Computation and Visualization
Version:	0.1.0
Description:	Provides a comprehensive, modular framework for computing the Soil Quality Index (SQI) using six established methods: Linear Scoring (Doran and Parkin, 1994, <doi:10.2136/sssaspecpub35.c1>), Regression-based (Masto et al., 2008, <doi:10.1007/s10661-007-9697-z>), Principal Component Analysis-based (Andrews et al., 2004, <doi:10.2136/sssaj2004.1945>), Fuzzy Logic, Entropy Weighting (Shannon, 1948, <doi:10.1002/j.1538-7305.1948.tb01338.x>), and TOPSIS (Hwang and Yoon, 1981, <doi:10.1007/978-3-642-48318-9>). Implements four variable scoring functions: more-is-better, less-is-better, optimum-value, and trapezoidal, following Karlen and Stott (1994, <doi:10.2136/sssaspecpub35.c4>). Includes automated Minimum Data Set selection via Principal Component Analysis with Variance Inflation Factor filtering (Kaiser, 1960, <doi:10.1177/001316446002000116>), one-way ANOVA with Tukey HSD post-hoc tests, leave-one-out sensitivity analysis, and publication-quality visualization using 'ggplot2'.
License:	GPL (≥ 3)
Encoding:	UTF-8
Language:	en-US
LazyData:	true
LazyDataCompression:	xz
RoxygenNote:	7.3.3
Depends:	R (≥ 4.0.0)
Imports:	dplyr (≥ 1.0.0), tidyr (≥ 1.1.0), ggplot2 (≥ 3.3.0), FactoMineR (≥ 2.4), factoextra (≥ 1.0.7), stats, methods, rlang (≥ 0.4.0), matrixStats (≥ 0.61.0), glmnet (≥ 4.1.0), car (≥ 3.0.0)
Suggests:	openxlsx (≥ 4.2.4), readxl (≥ 1.3.1), ggpubr (≥ 0.4.0), testthat (≥ 3.0.0), knitr (≥ 1.33), rmarkdown (≥ 2.11), spelling (≥ 2.2), covr (≥ 3.5.0)
VignetteBuilder:	knitr
Config/testthat/edition:	3
NeedsCompilation:	no
Packaged:	2026-04-13 10:00:29 UTC; acer
Author:	Sadikul Islam [aut, cre], Rajesh Kaushal [aut]
Maintainer:	Sadikul Islam <sadikul.islamiasri@gmail.com>
Repository:	CRAN
Date/Publication:	2026-04-20 12:30:02 UTC

SQIpro: Comprehensive Soil Quality Index Computation and Visualization

Description

Provides a comprehensive, modular framework for computing the Soil Quality Index (SQI) using six established methods: Linear Scoring (Doran and Parkin, 1994, doi:10.2136/sssaspecpub35.c1), Regression-based (Masto et al., 2008, doi:10.1007/s10661-007-9697-z), Principal Component Analysis-based (Andrews et al., 2004, doi:10.2136/sssaj2004.1945), Fuzzy Logic, Entropy Weighting (Shannon, 1948, doi:10.1002/j.1538-7305.1948.tb01338.x), and TOPSIS (Hwang and Yoon, 1981, doi:10.1007/978-3-642-48318-9). Implements four variable scoring functions: more-is-better, less-is-better, optimum-value, and trapezoidal, following Karlen and Stott (1994, doi:10.2136/sssaspecpub35.c4). Includes automated Minimum Data Set selection via Principal Component Analysis with Variance Inflation Factor filtering (Kaiser, 1960, doi:10.1177/001316446002000116), one-way ANOVA with Tukey HSD post-hoc tests, leave-one-out sensitivity analysis, and publication-quality visualization using 'ggplot2'.

Author(s)

Maintainer: Sadikul Islam sadikul.islamiasri@gmail.com (ORCID)

Authors:

• Rajesh Kaushal

Build a Variable Configuration Table

Description

Constructs a variable configuration data frame that specifies the scoring function type and relevant parameters for each soil indicator. This configuration table is the central object passed to all scoring and indexing functions in SQIpro.

Usage

make_config(
  variable,
  type,
  opt_low = rep(NA_real_, length(variable)),
  opt_high = rep(NA_real_, length(variable)),
  min_val = rep(NA_real_, length(variable)),
  max_val = rep(NA_real_, length(variable)),
  weight = rep(1, length(variable)),
  description = rep(NA_character_, length(variable))
)

Arguments

Value

A data frame (class sqi_config) with one row per variable.

References

Doran, J.W., & Parkin, T.B. (1994). Defining and assessing soil quality. In J.W. Doran et al. (Eds.), Defining Soil Quality for a Sustainable Environment, pp. 1–21. SSSA Special Publication 35. doi:10.2136/sssaspecpub35.c1

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Examples

cfg <- make_config(
  variable    = c("pH",  "EC",  "BD",  "OC",  "MBC", "Clay"),
  type        = c("opt", "less","less","more","more","opt"),
  opt_low     = c(6.0,   NA,    NA,    NA,    NA,    20),
  opt_high    = c(7.0,   NA,    NA,    NA,    NA,    35),
  description = c("Soil pH (H2O)",
                  "Electrical Conductivity (dS/m)",
                  "Bulk Density (g/cm3)",
                  "Organic Carbon (%)",
                  "Microbial Biomass Carbon (mg/kg)",
                  "Clay content (%)")
)
print(cfg)

PCA Biplot of Soil Variables and Groups

Description

Renders a PCA biplot showing both variable loadings and group centroids, using factoextra::fviz_pca_biplot. Useful for understanding variable relationships and group separation underlying MDS selection.

Usage

plot_pca_biplot(
  mds,
  scored,
  group_col = "LandUse",
  title = "PCA Biplot of Soil Quality Variables"
)

Arguments

Value

A ggplot object.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
mds    <- select_mds(scored, group_cols = c("LandUse","Depth"))
plot_pca_biplot(mds, scored, group_col = "LandUse")

Radar / Spider Chart of Variable Scores

Description

Draws a radar (spider) chart comparing mean variable scores across groups. Useful for visualising the multidimensional soil quality profile of each land-use system.

Usage

plot_radar(
  scored,
  config,
  group_col,
  group_cols = group_col,
  vars = NULL,
  title = "Soil Quality Radar Profile",
  palette = c("#1b7837", "#762a83", "#d6604d", "#4393c3", "#f4a582")
)

Arguments

Value

Invisible NULL; the chart is rendered to the active graphics device.

References

Chambers, J.M., & Hastie, T.J. (1992). Statistical Models in S. Wadsworth & Brooks/Cole.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
plot_radar(scored, cfg, group_col = "LandUse",
           group_cols = c("LandUse","Depth"))

Plot Individual Variable Scores as a Heatmap

Description

Displays a heatmap of mean variable scores (0–1) per group, allowing rapid visual identification of which variables drive high or low SQI within each land-use system.

Usage

plot_scores(
  scored,
  config,
  group_cols = "LandUse",
  group_by = group_cols[1],
  facet_by = NULL,
  palette = "RdYlGn",
  title = "Mean Variable Scores by Group"
)

Arguments

Value

A ggplot object.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
plot_scores(scored, cfg, group_cols = c("LandUse","Depth"),
            group_by = "LandUse", facet_by = "Depth")

Plot Scoring Curves for All Variables

Description

Visualises the scoring function (0–1 transformation) for each variable in the configuration, overlaid on the observed data distribution. This plot is essential for verifying that the scoring configuration is biologically sensible before computing indices.

Usage

plot_scoring_curves(data, config, group_cols = "LandUse", ncol = 3)

Arguments

Value

A ggplot object.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
plot_scoring_curves(soil_data, cfg,
                    group_cols = c("LandUse","Depth"))

Sensitivity Tornado Chart

Description

Visualises variable importance from sqi_sensitivity as a horizontal bar (tornado) chart, ordered from most to least sensitive.

Usage

plot_sensitivity(sa_result, title = "Variable Sensitivity to SQI")

Arguments

Value

A ggplot object.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
sa     <- sqi_sensitivity(scored, cfg, group_cols = c("LandUse","Depth"))
plot_sensitivity(sa)

Plot Soil Quality Index Across Groups

Description

Creates a grouped bar chart of SQI values per group, with optional error bars (standard deviation computed across replicate observations before indexing) and significance letters.

Usage

plot_sqi(
  sqi_result,
  sqi_col,
  group_col,
  fill_col = NULL,
  letters_df = NULL,
  title = "Soil Quality Index",
  y_label = "SQI (0-1)",
  palette = c("#2d6a4f", "#52b788", "#95d5b2", "#d8f3dc", "#b7e4c7")
)

Arguments

Value

A ggplot object.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
res    <- sqi_linear(scored, cfg, group_cols = c("LandUse","Depth"))
plot_sqi(res, sqi_col = "SQI_linear", group_col = "LandUse",
         fill_col = "Depth")

Score All Variables Using a Configuration Table

Description

Applies the appropriate scoring function to each soil variable according to a configuration table produced by make_config. This is the primary data-preparation step before computing any Soil Quality Index.

Usage

score_all(data, config, group_cols = "LandUse", custom_fns = list())

Arguments

Value

A data frame with the same structure as data, but with each variable column replaced by its 0–1 score. Group columns are preserved unchanged.

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH",  "EC",   "BD",   "OC",   "MBC",  "Clay"),
  type     = c("opt", "less", "less", "more", "more", "opt"),
  opt_low  = c(6.0,   NA,     NA,     NA,     NA,     20),
  opt_high = c(7.0,   NA,     NA,     NA,     NA,     35)
)
scored <- score_all(soil_data, cfg,
                    group_cols = c("LandUse", "Depth"))
head(scored)

Score a Variable With a User-Defined Function

Description

Applies an arbitrary user-defined scoring function to a numeric vector. The function must accept a numeric vector and return a numeric vector of the same length with values in [0, 1].

Usage

score_custom(x, FUN, ...)

Arguments

Value

Numeric vector of scores in [0, 1].

Examples

# Log-linear scoring for a skewed variable
mbc <- c(30, 80, 200, 400, 600)
score_custom(mbc, FUN = function(x) {
  s <- (log(x) - log(min(x))) / (log(max(x)) - log(min(x)))
  pmin(pmax(s, 0), 1)
})

Score a Variable Where Lower Values Are Better

Description

Applies a "less is better" linear scoring function, transforming raw variable values to a 0–1 score. Suitable for soil indicators where lower values denote better soil quality, such as bulk density, electrical conductivity, or heavy metal concentrations (Andrews et al., 2004).

The score is computed as:

S_i = \frac{x_{\max} - x_i}{x_{\max} - x_{\min}}

Usage

score_less(x, x_min = NULL, x_max = NULL)

Arguments

Value

Numeric vector of scores in [0, 1].

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Examples

bd <- c(0.9, 1.1, 1.3, 1.5, 1.7)   # Bulk Density (g/cm3)
score_less(bd)

# With domain bounds
score_less(bd, x_min = 0.8, x_max = 2.0)

Score a Variable Where Higher Values Are Better

Description

Applies a "more is better" linear scoring function, transforming raw variable values to a 0–1 score. This is appropriate for soil indicators where higher values improve soil function, such as organic carbon, microbial biomass, or cation exchange capacity (Andrews et al., 2004; Karlen & Stott, 1994).

The score is computed as:

S_i = \frac{x_i - x_{\min}}{x_{\max} - x_{\min}}

where x_{\min} and x_{\max} are taken from the observed data (or from user-supplied bounds).

Usage

score_more(x, x_min = NULL, x_max = NULL)

Arguments

Value

Numeric vector of scores in [0, 1].

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Karlen, D.L., & Stott, D.E. (1994). A framework for evaluating physical and chemical indicators of soil quality. In J.W. Doran et al. (Eds.), Defining Soil Quality for a Sustainable Environment, pp. 53–72. SSSA Special Publication 35. doi:10.2136/sssaspecpub35.c4

Examples

oc <- c(0.5, 1.2, 2.1, 3.4, 4.5)   # Organic Carbon (%)
score_more(oc)

# With user-defined bounds (e.g., 0 to 5%)
score_more(oc, x_min = 0, x_max = 5)

Score a Variable With an Optimum Value or Range (Bell Curve)

Description

Applies a bell-shaped (peaked) scoring function appropriate for soil variables that have an optimum range, beyond which both higher and lower values reduce soil quality. Classic examples include pH (optimal 6.0–7.0 for most crops) and clay content (Liebig et al., 1996; Karlen & Stott, 1994).

The scoring rules are:

• S = 1 if x \in [\code{opt\_low}, \code{opt\_high}]

• S = (x - x_{\min}) / (\code{opt\_low} - x_{\min}) if x < \code{opt\_low}

• S = (x_{\max} - x) / (x_{\max} - \code{opt\_high}) if x > \code{opt\_high}

Usage

score_optimum(x, opt_low, opt_high, x_min = NULL, x_max = NULL)

Arguments

Value

Numeric vector of scores in [0, 1].

References

Karlen, D.L., & Stott, D.E. (1994). A framework for evaluating physical and chemical indicators of soil quality. In J.W. Doran et al. (Eds.), Defining Soil Quality for a Sustainable Environment, pp. 53–72. SSSA Special Publication 35. doi:10.2136/sssaspecpub35.c4

Liebig, M.A., Varvel, G., & Doran, J.W. (1996). A simple performance- based index for assessing multiple agroecosystem functions. Agronomy Journal, 88, 739–745. doi:10.2134/agronj1996.00021962008800050011x

Examples

ph <- c(4.5, 5.5, 6.2, 6.8, 7.0, 7.5, 8.2)
score_optimum(ph, opt_low = 6.0, opt_high = 7.0)

clay <- c(10, 18, 25, 32, 45, 60)
score_optimum(clay, opt_low = 20, opt_high = 35)

Score a Variable With a Trapezoidal Function

Description

Applies a trapezoidal scoring function where scores are 1 within an ideal plateau [opt_low, opt_high], rise linearly from 0 at min_val to 1 at opt_low, and fall linearly from 1 at opt_high to 0 at max_val. Values outside [min_val, max_val] receive a score of 0.

This function is more flexible than score_optimum because the user explicitly controls the zero-score boundaries, making it suitable for variables with well-established critical thresholds.

Usage

score_trapezoid(x, min_val, opt_low, opt_high, max_val)

Arguments

Value

Numeric vector of scores in [0, 1].

References

Wymore, A.W. (1993). Model-Based Systems Engineering. CRC Press, Boca Raton, FL.

Buse, R., & Lele, S. (2003). Trapezoidal membership functions in fuzzy soil quality assessment. Geoderma, 114, 177–196.

Examples

ph <- c(3.5, 5.0, 6.5, 7.0, 7.8, 8.5, 9.5)
# pH: absolute zero below 4 and above 9; ideal 6.0-7.0
score_trapezoid(ph, min_val = 4.0, opt_low = 6.0,
                opt_high = 7.0, max_val = 9.0)

Select a Minimum Data Set (MDS) of Soil Quality Indicators

Description

Identifies the most informative subset of soil variables (the Minimum Data Set, MDS) using Principal Component Analysis (PCA). Only variables with high factor loadings on principal components explaining eigenvalue > 1 (Kaiser criterion) are retained. Where multiple variables load highly on the same component, the one with the highest correlation to others in that component is selected to minimise redundancy.

This approach follows the widely cited method of Andrews et al. (2004) and Sharma et al. (2008), and is equivalent to the PCAIndex algorithm in Wani et al. (2023).

Usage

select_mds(
  data,
  group_cols = "LandUse",
  load_threshold = 0.5,
  vif_threshold = 10,
  n_pc = "auto",
  verbose = TRUE
)

Arguments

Details

**Algorithm steps:**

1. Standardise all numeric variables (mean = 0, sd = 1).

2. Perform PCA; retain components with eigenvalue > 1.

3. For each retained component, identify variables with absolute loading \ge load_threshold.

4. Among those, select the variable with the highest sum of absolute Pearson correlations to all others in the set (i.e., the most correlated, least redundant variable).

5. Optionally, remove variables with high Variance Inflation Factor (VIF > vif_threshold) among the MDS candidates.

Value

A list of class sqi_mds with:

mds_vars

Character vector of selected MDS variable names.

all_vars

Character vector of all candidate variable names.

pca

The PCA result object.

loadings

Matrix of factor loadings.

eigenvalues

Numeric vector of eigenvalues.

var_explained

Numeric vector of variance explained (%) per component.

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework: A quantitative soil quality evaluation method. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Kaiser, H.F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20(1), 141–151. doi:10.1177/001316446002000116

Sharma, K.L., et al. (2008). Long-term soil management effects on soil quality indices. Geoderma, 144, 290–300. doi:10.1016/j.geoderma.2007.11.019

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","PMN","Clay","WHC","DEH","AP","TN"),
  type     = c("opt","less","less","more","more","more",
               "opt","more","more","more","more"),
  opt_low  = c(6.0, NA, NA, NA, NA, NA, 20, NA, NA, NA, NA),
  opt_high = c(7.0, NA, NA, NA, NA, NA, 35, NA, NA, NA, NA)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
mds    <- select_mds(scored, group_cols = c("LandUse","Depth"))
mds$mds_vars

Hypothetical Soil Quality Dataset

Description

A hypothetical dataset representing soil physicochemical and biological properties across five land-use systems and two soil depths, generated using realistic parameter ranges reported in the soil quality literature. This dataset is intended for demonstrating the functions in the SQIpro package and for pedagogical purposes.

Usage

soil_data

Format

A data frame with 100 rows and 14 variables:

LandUse

Character. Land-use system: Natural_Forest, Agroforestry, Cropland, Grassland, Degraded_Land.

Depth

Character. Soil depth: Surface_0_15cm or Subsurface_15_30cm.

pH

Numeric. Soil pH (1:2.5 water suspension). Unitless. Optimal range 6.0–7.0 for most crops (Brady & Weil, 2008).

EC

Numeric. Electrical Conductivity (dS m^{-1}). Lower values indicate less salinity stress; <0.2 dS m^{-1} considered non-saline (Richards, 1954).

BD

Numeric. Bulk Density (g cm^{-3}). Lower values indicate better soil structure and aeration; >1.6 g cm^{-3} restricts root growth (Arshad et al., 1996).

CEC

Numeric. Cation Exchange Capacity (cmol(+) kg^{-1}). Higher values indicate greater nutrient-holding capacity (Sparks, 2003).

OC

Numeric. Organic Carbon (%). Higher values indicate greater soil organic matter, a key indicator of soil health (Doran & Parkin, 1994).

MBC

Numeric. Microbial Biomass Carbon (mg kg^{-1}). Indicator of soil biological activity (Brookes, 1995).

PMN

Numeric. Potentially Mineralizable Nitrogen (mg kg^{-1}). Indicates N-supplying capacity of soil (Stanford & Smith, 1972).

Clay

Numeric. Clay content (%). Optimal range 20–35% for water and nutrient retention (Arshad et al., 1996).

WHC

Numeric. Water Holding Capacity (%). Higher values indicate better moisture retention (Reynolds et al., 2009).

DEH

Numeric. Dehydrogenase Enzyme Activity (\mug TPF g^{-1} day^{-1}). Indicator of overall microbial metabolic activity (Casida et al., 1964).

AP

Numeric. Available Phosphorus (mg kg^{-1}). Higher values indicate better P availability for plants (Olsen & Sommers, 1982).

TN

Numeric. Total Nitrogen (%). Higher values indicate greater N reserves (Bremner, 1996).

Details

Parameter ranges were informed by values reported in:

• Doran and Parkin (1994) for biological indicators

• Andrews et al. (2004) for MDS indicator ranges

• Masto et al. (2008) for land-use comparison ranges

The dataset is entirely synthetic and does not represent any specific geographic location.

Source

Synthetically generated for the SQIpro package.

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework: A quantitative soil quality evaluation method. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Arshad, M.A., Lowery, B., & Grossman, B. (1996). Physical tests for monitoring soil quality. In J.W. Doran & A.J. Jones (Eds.), Methods for Assessing Soil Quality, pp. 123–141. SSSA Special Publication 49. doi:10.2136/sssaspecpub49.c7

Brady, N.C., & Weil, R.R. (2008). The Nature and Properties of Soils (14th ed.). Prentice Hall, New Jersey.

Doran, J.W., & Parkin, T.B. (1994). Defining and assessing soil quality. In J.W. Doran et al. (Eds.), Defining Soil Quality for a Sustainable Environment, pp. 1–21. SSSA Special Publication 35. doi:10.2136/sssaspecpub35.c1

Masto, R.E., Chhonkar, P.K., Singh, D., & Patra, A.K. (2008). Alternative soil quality indices for evaluating the effect of intensive cropping, fertilisation and manuring for 31 years in the semi-arid soils of India. Environmental Monitoring and Assessment, 136, 419–435. doi:10.1007/s10661-007-9697-z

Examples

data(soil_data)
head(soil_data)
summary(soil_data)
table(soil_data$LandUse, soil_data$Depth)

One-Way ANOVA and Tukey HSD Post-Hoc Test for SQI

Description

Performs a one-way ANOVA to test whether Soil Quality Index values differ significantly across land-use groups, followed by Tukey's Honest Significant Difference (HSD) test for pairwise comparisons.

Usage

sqi_anova(scored, sqi_col, group_col, alpha = 0.05)

Arguments

Value

A list with:

anova_table

An anova object.

tukey

A TukeyHSD object.

significant

Logical. Whether the ANOVA is significant at alpha.

compact_letters

Data frame of compact letter display for plotting.

References

Tukey, J.W. (1949). Comparing individual means in the analysis of variance. Biometrics, 5(2), 99–114. doi:10.2307/3001913

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
# Compute per-observation linear SQI for ANOVA
scored$SQI_obs <- rowMeans(scored[, cfg$variable], na.rm = TRUE)
aov_result <- sqi_anova(scored, sqi_col = "SQI_obs",
                         group_col = "LandUse")
print(aov_result$tukey)

Compare All SQI Methods

Description

Runs all six SQI methods (sqi_linear, sqi_regression, sqi_pca, sqi_fuzzy, sqi_entropy, sqi_topsis) on the same scored dataset and returns a combined results table for method comparison.

Usage

sqi_compare(scored, config, group_cols = "LandUse", dep_var = NULL, mds = NULL)

Arguments

Value

A data frame with one row per group and columns for each SQI method. Also includes Mean_SQI and Rank columns.

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored  <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
results <- sqi_compare(scored, cfg, group_cols = c("LandUse","Depth"),
                        dep_var = "OC")
print(results)

Soil Quality Index: Entropy Weighting Method

Description

Computes SQI using Shannon entropy to derive objective weights for each variable. Variables with higher information entropy (greater discriminating power among groups) receive higher weights. This removes subjectivity from weight assignment.

The entropy weight for variable j is:

e_j = -\frac{1}{\ln n} \sum_{i=1}^{n} p_{ij} \ln(p_{ij})

w_j = \frac{1 - e_j}{\sum_k (1 - e_k)}

where p_{ij} = \bar{S}_{ij} / \sum_i \bar{S}_{ij}.

Usage

sqi_entropy(scored, config, group_cols = "LandUse", mds_vars = NULL)

Arguments

Value

A data frame with group columns, SQI_entropy, and attribute entropy_weights (named numeric vector).

References

Shannon, C.E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27(3), 379–423. doi:10.1002/j.1538-7305.1948.tb01338.x

Li, P., Qian, H., & Wu, J. (2010). Groundwater quality assessment based on improved water quality index in Pengyang County, Ningxia, Northwest China. E-Journal of Chemistry, 7, 209–216. doi:10.1155/2010/451304

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
result <- sqi_entropy(scored, cfg, group_cols = c("LandUse","Depth"))
attr(result, "entropy_weights")

Soil Quality Index: Fuzzy Logic Method

Description

Computes SQI using a fuzzy membership aggregation approach. Each scored variable (already 0–1) is treated as a fuzzy membership value, and groups are aggregated using either the arithmetic mean (equivalent to the linear method) or the fuzzy weighted average operator.

This approach is appropriate when variable importance is uncertain or when expert-elicited weights are available (Zhu et al., 2006; Torbert & Wood, 1992).

Usage

sqi_fuzzy(
  scored,
  config,
  group_cols = "LandUse",
  mds_vars = NULL,
  fuzzy_weights = NULL,
  operator = c("mean", "geometric")
)

Arguments

Value

A data frame with group columns and SQI_fuzzy (0–1).

References

Zhu, A.X., Liu, F., Li, B., Pei, T., Qin, C., Liu, G., Wang, Y., Chen, Y., Ma, X., Qi, F., & Li, R. (2010). Differentiation of soil conditions over flat areas using land surface feedback dynamic patterns extracted from MODIS. Soil Science Society of America Journal, 74(1), 861–869.

Torbert, H.A., & Wood, C.W. (1992). Effects of soil compaction and water-filled pore space on soil microbial activity and N losses. Communications in Soil Science and Plant Analysis, 23, 1321–1331. doi:10.1080/00103629209368668

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
result <- sqi_fuzzy(scored, cfg, group_cols = c("LandUse","Depth"))
print(result)

Soil Quality Index: Linear Scoring Method

Description

Computes the Soil Quality Index (SQI) using the linear additive scoring method of Doran & Parkin (1994) and Andrews et al. (2004). Each variable score (0–1, from score_all) is averaged across replicates within each group, optionally weighted, and then min-max normalised to produce the final index.

SQI_g = \frac{\sum_{j=1}^{p} w_j \bar{S}_{gj}}{\sum_{j=1}^{p} w_j}

where \bar{S}_{gj} is the mean score of variable j in group g and w_j is the weight of variable j.

Usage

sqi_linear(
  scored,
  config,
  group_cols = "LandUse",
  mds_vars = NULL,
  weights = NULL
)

Arguments

Value

A data frame with group columns plus:

SQI_linear

Final normalised Soil Quality Index (0–1).

Raw_score

Weighted mean score before normalisation.

References

Doran, J.W., & Parkin, T.B. (1994). Defining and assessing soil quality. In J.W. Doran et al. (Eds.), Defining Soil Quality for a Sustainable Environment, pp. 1–21. SSSA Special Publication 35. doi:10.2136/sssaspecpub35.c1

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
result <- sqi_linear(scored, cfg, group_cols = c("LandUse","Depth"))
print(result)

Soil Quality Index: PCA-Based Method

Description

Computes SQI using Principal Component Analysis, weighting selected MDS variables by the proportion of variance their component explains. This is the most widely cited data-driven approach in soil quality research (Andrews et al., 2004; Bastida et al., 2008).

SQI_{PCA} = \sum_{k=1}^{m} \frac{V_k}{\sum V} \bar{S}_{g, j_k}

where V_k is the variance explained by component k, j_k is the MDS variable selected from component k, and \bar{S}_{g, j_k} is the group mean score of that variable.

Usage

sqi_pca(scored, config, group_cols = "LandUse", mds = NULL)

Arguments

Value

A data frame with group columns and SQI_pca (0–1).

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Bastida, F., Zsolnay, A., Hernandez, T., & Garcia, C. (2008). Past, present and future of soil quality indices: A biological perspective. Geoderma, 147(3–4), 159–171. doi:10.1016/j.geoderma.2008.08.007

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
result <- sqi_pca(scored, cfg, group_cols = c("LandUse","Depth"))
print(result)

Soil Quality Index: Regression-Based Method

Description

Computes the SQI using stepwise multiple linear regression to identify and weight the most predictive soil variables. The dependent variable (e.g., crop yield, total biomass) determines which variables enter the model. Regression coefficients serve as weights in the index.

This follows the method described by Masto et al. (2008) and Mukherjee & Lal (2014).

Usage

sqi_regression(
  scored,
  config,
  dep_var,
  group_cols = "LandUse",
  mds_vars = NULL,
  direction = "both"
)

Arguments

Value

A data frame with group columns plus:

SQI_regression

Normalised SQI (0–1).

selected_vars

(attribute) Character vector of selected predictors.

References

Masto, R.E., Chhonkar, P.K., Singh, D., & Patra, A.K. (2008). Alternative soil quality indices. Environmental Monitoring and Assessment, 136, 419–435. doi:10.1007/s10661-007-9697-z

Mukherjee, A., & Lal, R. (2014). Comparison of soil quality index using three methods. PLOS ONE, 9(8), e105981. doi:10.1371/journal.pone.0105981

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
# OC used as surrogate dependent variable
result <- sqi_regression(scored, cfg, dep_var = "OC",
                         group_cols = c("LandUse","Depth"))
print(result)

Sensitivity Analysis for SQI

Description

Quantifies the contribution of each soil variable to the overall Soil Quality Index by a leave-one-out approach: each variable is removed in turn and the resulting index is compared to the full index. A larger change indicates a higher-sensitivity (more important) variable.

Usage

sqi_sensitivity(
  scored,
  config,
  group_cols = "LandUse",
  method = c("linear", "fuzzy", "entropy", "topsis"),
  mds_vars = NULL
)

Arguments

Value

A data frame with columns variable, mean_change (mean absolute change in SQI when variable is removed), sd_change, and relative_importance (0–1, normalised).

References

Saltelli, A., Ratto, M., Andres, T., Campolongo, F., Cariboni, J., Gatelli, D., Saisana, M., & Tarantola, S. (2008). Global Sensitivity Analysis: The Primer. John Wiley & Sons, Chichester. doi:10.1002/9780470725184

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored  <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
sa      <- sqi_sensitivity(scored, cfg, group_cols = c("LandUse","Depth"))
print(sa)

Soil Quality Index: TOPSIS Method

Description

Computes SQI using the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), a multi-criteria decision analysis method. Each group is ranked by its Euclidean distance to the positive ideal solution (all scores = 1) and negative ideal solution (all scores = 0).

C_i^* = \frac{d_i^-}{d_i^+ + d_i^-}

where d_i^+ and d_i^- are distances to the positive and negative ideal solutions. C_i^* \in [0, 1] with higher values indicating better soil quality.

Usage

sqi_topsis(
  scored,
  config,
  group_cols = "LandUse",
  mds_vars = NULL,
  weights = NULL
)

Arguments

Value

A data frame with group columns and SQI_topsis (0–1).

References

Hwang, C.L., & Yoon, K. (1981). Multiple Attribute Decision Making: Methods and Applications. Springer, Berlin. doi:10.1007/978-3-642-48318-9

Yoon, K. (1987). A reconciliation among discrete compromise solutions. Journal of the Operational Research Society, 38, 277–286. doi:10.1057/jors.1987.44

Examples

data(soil_data)
cfg <- make_config(
  variable = c("pH","EC","BD","OC","MBC","Clay"),
  type     = c("opt","less","less","more","more","opt"),
  opt_low  = c(6.0, NA, NA, NA, NA, 20),
  opt_high = c(7.0, NA, NA, NA, NA, 35)
)
scored <- score_all(soil_data, cfg, group_cols = c("LandUse","Depth"))
result <- sqi_topsis(scored, cfg, group_cols = c("LandUse","Depth"))
print(result)

Validate Input Data for SQI Analysis

Description

Checks that a data frame meets requirements for Soil Quality Index (SQI) computation: correct column types, sufficient sample sizes, absence of infinite values, and appropriate variable configuration.

Usage

validate_data(
  data,
  group_cols = NULL,
  config = NULL,
  min_n = 3,
  verbose = TRUE
)

Arguments

Value

Invisibly returns a list with components:

valid

Logical. TRUE if all checks pass.

messages

Character vector of warning/info messages.

n_per_group

Data frame of group sizes.

References

Andrews, S.S., Karlen, D.L., & Cambardella, C.A. (2004). The soil management assessment framework: A quantitative soil quality evaluation method. Soil Science Society of America Journal, 68(6), 1945–1962. doi:10.2136/sssaj2004.1945

Examples

data(soil_data)
result <- validate_data(soil_data, group_cols = c("LandUse", "Depth"))
result$valid
result$n_per_group