Skip to contents

Comprehensive Analysis of Variance (ANOVA) Engine for Completely Randomized Design (CRD)

anova_crd.Rd

The `anova_crd` function executes a complete, high-precision linear model analysis for agricultural, laboratory, or greenhouse trials laid out under a Completely Randomized Design (CRD). It computes partition sums of squares, hypothesis testing statistics, treatment variances, significance flags, and the Coefficient of Variation (CV

Usage

anova_crd(data, trait, reporting_level = 2)

Arguments

data

A verified data.frame containing the columns Genotype and the target phenotypic trait response.

trait

A single character string specifying the exact column name of the numeric trait to analyze.

reporting_level

An integer flag defining console trace settings: 0 for silent execution, 1 for printing basic ANOVA summary tables, and 2 for comprehensive diagnostic trace logs. Defaults to 2.

Value

Invisibly returns a structured named list of class "list" containing 4 computational components:

anova_table

A data.frame acting as the standard ANOVA source matrix table for CRD, containing degrees of freedom, sums of squares, mean squares, F-statistics, and p-values.

cv_percentage

A numeric scalar representing the computed Coefficient of Variation percentage (CV%).

mean_square_error

A numeric scalar representing the isolated Residual Error Mean Square (EMS), ready for downstream evaluation.

grand_mean

A numeric scalar representing the overall arithmetic mean value of the evaluated phenotypic trait.

If reporting_level >= 1, the compiled ANOVA table along with the grand mean and CV% are printed directly to the console before returning the list.

Details

In laboratory experiments, growth chamber studies, or field trials with completely homogeneous environments, blocking is unnecessary. This function utilizes standard least-squares projection to build the classic orthogonal CRD ANOVA matrix, modeling the response vector as a function of treatment effects without blocking constraints: $$Y_{ij} = \mu + T_i + \varepsilon_{ij}$$ Where \(T_i\) represents the treatment/genotype effect, and \(\varepsilon_{ij}\) is the residual experimental error. The function handles both balanced and unbalanced data structures perfectly, ensuring proper adjustments to degrees of freedom if replication numbers vary across lines.

See also

anova_rcbd

Examples

   # Execute complete CRD partition on your actual target trait (GYPM)
   crd_results <- anova_crd(data = gv_data, trait = "GYPM")
#> =====================================================================================
#> AGRIDATATOOLS PACKAGED ENGINE v0.1.0 - COMPLETELY RANDOMIZED DESIGN (CRD) ANOVA
#> Analysis Inception:  2026-08-17 12:02:15.203653 
#> -------------------------------------------------------------------------------------
#> [DIAGNOSTIC - MODEL]: Fitting one-way orthogonal linear model matrix formulas...
#> 
#> ---------------------------------------------------------------------------
#>  ANALYSIS OF VARIANCE (ANOVA) FOR CRD - TRAIT:  GYPM 
#> ---------------------------------------------------------------------------
#>                  Source  Df        SS        MS F_value p_value
#>  Genotypes (Treatments)  39 73217.300 1877.3667 109.895 < 0.001
#>        Error (Residual)  80  1366.667   17.0833      NA      NA
#>                   Total 119 74583.967        NA      NA      NA
#> ---------------------------------------------------------------------------
#>  Trial Grand Mean                 :  104.9833 
#>  Coefficient of Variation (CV%)   :  3.94 %
#> =====================================================================================
#> 
#> [LOG - FINALIZE]: CRD matrix compilation finished in  0.00539  seconds.