Cleaning Validation Package

Description

This package offers a comprehensive suite of functions for cleaning validation, a critical component of quality control in pharmaceutical manufacturing. The included functions assist in analyzing residue data, evaluating cleaning efficacy, and ensuring that cleaning processes meet regulatory standards.

Details

The functions primarily return data frames, streamlining data preprocessing, analysis, and the application of statistical methods for cleaning process evaluation. This toolset simplifies the workflow for cleaning validation professionals, providing resources for various tasks. Function cv01 cleans three data types. Functions cv02 to cv12 (excluding cv05) are designed for sequential DAR and CAR analysis. Functions cv13 and cv14 assess whether Mic follows a Poisson distribution. For Mic data that follows a Poisson distribution, function cv05 and functions cv15 to cv29 should be used in sequence. If Mic data do not follow a Poisson distribution, function cv05 and functions cv02 to cv12 (excluding cv06) are applicable. Function cv30 synthesizes the Process Performance Index (Ppu) for DAR, CAR, and Mic. Supplementary to its core capabilities, the package includes datasets—Eq_DAR, Eq_CAR, and Eq_Mic—for demonstrating functionality in practical contexts.

License

This package is free software; you may redistribute and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License or (at your option) any later version.

Examples

## Not run: 
# Example code here to demonstrate package usage:
# This could include data loading, transforming, and cleaning validation analysis.

## End(Not run)

Equipment Cleaning Data for CAR

Description

A dataset containing cleaning validation data for equipment CAR.

Usage

Eq_CAR

Format

A data frame with 30 rows and 3 variables:

CAR

Numeric vector with CAR measurements.

USL

Numeric vector with Upper Specification Limits for CAR.

CleaningEvent

Factor vector with Cleaning Event identifiers.

Classification

Character vector with the deviation status for each cleaning event. Defaults to "normal".

LIMSProjectID

Integer or character vector with unique project IDs assigned to each row.

Source

Details about the data source.

Equipment Cleaning Data for DAR

Description

A dataset containing cleaning validation data for equipment DAR.

Usage

Eq_DAR

Format

A data frame with 60 rows and 3 variables:

DAR

Numeric vector with DAR measurements.

USL

Numeric vector with Upper Specification Limits for DAR.

CleaningEvent

Factor vector with Cleaning Event identifiers.

Classification

Character vector with the deviation status for each cleaning event. Defaults to "normal".

LIMSProjectID

Integer or character vector with unique project IDs assigned to each row.

Source

Details about the data source.

Equipment Cleaning Data for Microbial Bioburden

Description

A dataset containing cleaning validation data for microbial bioburden (Mic).

Usage

Eq_Mic

Format

A data frame with 20 rows and 3 variables:

Mic

Numeric vector with Mic measurements.

USL

Numeric vector with Upper Specification Limits for Mic.

CleaningEvent

Factor vector with Cleaning Event identifiers.

Classification

Character vector with the deviation status for each cleaning event. Defaults to "normal".

LIMSProjectID

Integer or character vector with unique project IDs assigned to each row.

Source

Details about the data source.

Clean and preprocess residue data for stability and capability analysis

Description

This function ensures data type and no missing data in residue_col, cleaning_event_col, usl_col of data their type. Furthermore, it changes cleaning_event_col to time ordered factor. It cleans and pre-processes the residue data for stability and capability analysis, ensuring that it meets the necessary criteria for analysis.

Usage

cv01_dfclean(data, residue_col, cleaning_event_col, usl_col)

Arguments

Value

A cleaned and pre-processed data frame such that all variables have no missing values, its CleaningEvent is time-ordered categorical variable, and Residue and USL are numeric.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assume Eq_DAR, Eq_CAR, and Eq_Mic are loaded datasets

# Clean and preprocess residue data for Eq_DAR
Eq_DAR <- cv01_dfclean(data = Eq_DAR, residue_col = "DAR", usl_col = "USL", 
cleaning_event_col = "CleaningEvent")

# Clean and preprocess residue data for Eq_CAR
Eq_CAR <- cv01_dfclean(data = Eq_CAR, residue_col = "CAR", usl_col = "USL", 
cleaning_event_col = "CleaningEvent")

# Clean and preprocess residue data for Eq_Mic
Eq_Mic <- cv01_dfclean(data = Eq_Mic, residue_col = "Mic", usl_col = "USL", 
cleaning_event_col = "CleaningEvent")

Summarize Non-Process Related OOS and Reswab Data Which May Not Be Included in the Analysis

Description

This function processes three datasets to identify unique project IDs based on non-process-related out-of-specification (OOS) and reswab cases, then summarizes this information into a dataframe. If your data does not have reswab or OOS, you do not need to use this function.

Usage

cv02_nonpro_oos_reswab(Eq_DAR, Eq_CAR, Eq_Mic)

Arguments

Value

A dataframe summarizing the non-process-related OOS and reswab data.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv02_nonpro_oos_reswab(Eq_DAR, Eq_CAR, Eq_Mic)

Unify USL Percentages for Specified Residue

Description

This function takes a dataset and computes the percentage of residue over USL for each event, as well as mean and median of these percentages for each cleaning event and overall.

Usage

cv03_usl_unification(data, cleaning_event_col, residue_col, usl_col)

Arguments

Value

A dataframe with original data and additional columns for residue percentages, and their mean and median values per cleaning event and overall.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv03_usl_unification(data = Eq_DAR, cleaning_event_col = "CleaningEvent", 
residue_col = "DAR", usl_col = "USL")

Plot Histogram with Kernel Density Estimate Curve

Description

This function takes a dataset and a column representing the residue percentages and generates a histogram overlaid with a KDE (Kernel Density Estimate) curve. It calculates and marks quantiles P0.5, P0.8413, P0.9772, and the P0.99865, i.e., UCL (Upper Control Limit) on the plot.

Usage

cv04_histogram_kde(data, residue_pct_col)

Arguments

Value

A ggplot object representing the histogram with KDE curve.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

Eq_DAR <- cv03_usl_unification(data=Eq_DAR,"CleaningEvent", "DAR", usl_col="USL")
cv04_histogram_kde(data = Eq_DAR, residue_pct_col = "DAR_Pct")

Perform Shapiro-Wilk Normality Test

Description

This function performs the Shapiro-Wilk test for normality on a specified variable in a dataset. It returns a data frame with the variable name, the Shapiro-Wilk statistic, the p-value in scientific notation, and an indication of whether the p-value is less than 0.05.

Usage

cv05_sw_norm_test_1(data, residue_col)

Arguments

Value

A data frame with the Shapiro-Wilk test results.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

data(Eq_Mic)
cv05_sw_norm_test_1(data=Eq_Mic, residue_col="Mic")

Perform Shapiro-Wilk Normality Test on Two Variables

Description

This function performs the Shapiro-Wilk test for normality on two specified variables within a dataset. It returns a data frame with the variables' names, Shapiro-Wilk statistics, p-values in scientific notation, and indications of whether the p-values are less than 0.05.

Usage

cv06_sw_norm_test_2(data, residue_col, residue_pct_col)

Arguments

Value

A data frame with Shapiro-Wilk test results for both variables.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# assuming Eq_DAR is a predefined dataset 
Eq_DAR <- cv03_usl_unification(data = Eq_DAR,  "CleaningEvent",  "DAR",  "USL")
cv06_sw_norm_test_2(data=Eq_DAR, residue_col="DAR", residue_pct_col="DAR_Pct")

Median Control Chart

Description

This function creates a control chart for the median residue percentages based on kernel density estimation. The in put residue_pct_meidan_col can be median of non-USL_unified variable such as Mic_Median, DAR_Median, or CAR_Median

Usage

cv071_median_control_chart(data, cleaning_event_col, residue_pct_median_col)

Arguments

Value

The meidan control chart.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_DAR' is a data frame with appropriate columns:
Eq_DAR <- cv03_usl_unification(data = Eq_DAR,  "CleaningEvent",  "DAR",  "USL")
cv07_median_control_chart(data = Eq_DAR, cleaning_event_col = "CleaningEvent", 
residue_pct_median_col="DAR_Pct_Median")

Median Control Chart and Density Plot

Description

This function creates a control chart and a density plot for the median residue percentages based on kernel density estimation.

Usage

cv07_median_control_chart(data, cleaning_event_col, residue_pct_median_col)

Arguments

Value

A cowplot object containing the combined control chart and density plot.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_DAR' is a data frame with appropriate columns:
Eq_DAR <- cv03_usl_unification(data = Eq_DAR,  "CleaningEvent",  "DAR",  "USL")
cv07_median_control_chart(data = Eq_DAR, cleaning_event_col = "CleaningEvent", 
residue_pct_median_col="DAR_Pct_Median")

Variability Chart for Cleaning Events

Description

This function generates a variability chart for cleaning events, showing data points, outliers, and overall statistics like the grand mean and median.

Usage

cv08_variability_chart(data, cleaning_event_col, residue_pct_col, usl_pct_col)

Arguments

Value

A ggplot object representing the variability chart.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

Eq_DAR <- cv01_dfclean(data=Eq_DAR, cleaning_event_col="CleaningEvent", 
residue_col="DAR", usl_col="USL" ) 
Eq_DAR <- cv03_usl_unification(data=Eq_DAR, cleaning_event_col="CleaningEvent", 
residue_col="DAR", usl_col="USL")
cv08_variability_chart(data=Eq_DAR, cleaning_event_col="CleaningEvent", 
residue_pct_col="DAR_Pct", usl_pct_col="USL_Pct")

Kruskal-Wallis Test for Residue Percentages

Description

Perform Kruskal-Wallis test for residue percentages based on cleaning events.

Usage

cv09_kw_test(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame of Kruskal-Wallis test results.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_DAR' is the data frame, 'DAR_Pct' is the residue column, 
# and 'CleaningEvent' is the cleaning event column.
Eq_DAR <- cv01_dfclean(data=Eq_DAR, cleaning_event_col="CleaningEvent", 
residue_col="DAR", usl_col="USL" ) 
Eq_DAR <- cv03_usl_unification(data=Eq_DAR, cleaning_event_col="CleaningEvent", 
residue_col="DAR", usl_col="USL")
kw_test_results <- cv09_kw_test(data = Eq_DAR, residue_col = "DAR_Pct", 
 cleaning_event_col = "CleaningEvent")

Dunn's Test for Residue

Description

Perform Dunn's test for residue based on cleaning events. Choose the control group as the cleaning event whose median is closest to the grand median. This function is for investigation purpose.

Usage

cv10_dunn_test_vs_control(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame of Dunn's test results with control group.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# 'Eq_DAR' is the  data frame, 'DAR_Pct' is the residue column, and
# 'CleaningEvent' is the cleaning event column.
Eq_DAR <- cv03_usl_unification(data = Eq_DAR, residue_col = "DAR",
cleaning_event_col = "CleaningEvent",  usl_col = "USL")
dunn_test_results_vs_control <- cv10_dunn_test_vs_control(data = Eq_DAR,
residue_col = "DAR_Pct", cleaning_event_col = "CleaningEvent")

Variability Components Analysis by Median with Bootstrap

Description

Perform Variability Components Analysis (VCA) by median for residue percentages based on cleaning events with bootstrap for confidence intervals.

Usage

cv11_vca_by_median(data, residue_col, cleaning_event_col, n_bootstrap = 2000)

Arguments

Value

A data frame summarizing variability components analysis by median along with confidence intervals from bootstrap.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_DAR' is the data frame, 'DAR_Pct' is the residue column, 
# and 'CleaningEvent' is the cleaning event column.
Eq_DAR <- cv01_dfclean(data=Eq_DAR, cleaning_event_col="CleaningEvent",
residue_col="DAR", usl_col="USL" ) 
Eq_DAR <- cv03_usl_unification(data=Eq_DAR, cleaning_event_col="CleaningEvent", 
residue_col="DAR", usl_col="USL")
summary <- cv11_vca_by_median(data = Eq_DAR, residue_col = "DAR_Pct", 
cleaning_event_col = "CleaningEvent", n_bootstrap = 2000)

Calculate PPU using KDE density estimation

Description

Calculate PPU using KDE density estimation

Usage

cv12_kde_ppu(
  data,
  residue_col,
  cleaning_event_col,
  usl_col,
  n_bootstrap = 1000
)

Arguments

Value

A dataframe with the estimated PPU and its 95

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

 Eq_DAR <- cv03_usl_unification(data = Eq_DAR, cleaning_event_col = "CleaningEvent", 
residue_col = "DAR", usl_col = "USL")
cv12_kde_ppu(data = Eq_DAR, residue_col = "DAR_Pct", cleaning_event_col = "CleaningEvent",
 usl_col = "USL_Pct", n_bootstrap = 1000)

Poisson Goodness-of-Fit Test

Description

Conducts a goodness-of-fit test to evaluate if the Mic data follows a Poisson distribution.

Usage

cv13_poisson_test(data, residue_col)

Arguments

Value

A dataframe object representing the chi-squared statistic and the p-value from the goodness-of-fit test.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming Eq_Mic is your dataframe and Mic is the column to be tested
 cv13_poisson_test(data=Eq_Mic, residue_col="Mic")

Dispersion Test for Poisson Regression Models

Description

Performs a dispersion test on a Poisson regression model to check for overdispersion. The function fits a Poisson regression model to the data using the specified columns, and then performs a dispersion test using the model.

Usage

cv14_dispersion_test(data, residue_col, cleaning_event_col)

Arguments

Value

A dataframe object with the results of the overdispersion test, including

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com] the Z-value, P-value, and dispersion estimate.

Examples

cv14_dispersion_test(data = Eq_Mic, residue_col = "Mic", cleaning_event_col = "CleaningEvent")

Calculate Mic Statistics

Description

Calculate Mic Statistics

Usage

cv15_mic_mutate(data, cleaning_event_col, residue_col)

Arguments

Value

A dataframe with new columns for mean, median, grand mean, and grand median of Mic values.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv15_mic_mutate(data=Eq_Mic, cleaning_event_col="CleaningEvent", residue_col="Mic")

Create a u-Chart for Poisson-distributed Data

Description

This function generates a u-chart for visualizing the stability and capability of a process based on a Poisson distribution.

Usage

cv16_u_chart(data, residue_col, cleaning_event_col)

Arguments

Value

A ggplot object representing the u-chart.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv16_u_chart(data = Eq_Mic, residue_col = "Mic", cleaning_event_col = "CleaningEvent")

Create a CUSUM Chart for Poisson-distributed Data

Description

This function computes the cumulative sum (CUSUM) for the mean values of a specified residue column aggregated by a cleaning event column. It then generates a CUSUM chart for visualizing the stability of a process based on a Poisson distribution. The reference value 'k' can be provided; if not, it defaults to half of the process average lambda.

Usage

cv17_cusum(data, residue_col, cleaning_event_col, k = NULL)

Arguments

Value

A ggplot object representing the CUSUM chart.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# To create a CUSUM chart with default k value
cv17_cusum(data = Eq_Mic, residue_col = "Mic", cleaning_event_col = "CleaningEvent")

# To create a CUSUM chart with a specified k value
cv17_cusum(data = Eq_Mic, residue_col = "Mic", cleaning_event_col = "CleaningEvent", k = 0.75)

Exponentially Weighted Moving Average (EWMA) Chart

Description

Generates an EWMA chart for a specified residue column grouped by cleaning events.

Usage

cv18_ewma(data, residue_col, cleaning_event_col, alpha = 0.2)

Arguments

Value

A ggplot object representing the EWMA chart.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_Mic' is a data frame, 'Mic' is the residue column of interest,
# and 'CleaningEvent' is the column representing cleaning events.
ewma_plot <- cv18_ewma(data = Eq_Mic, residue_col = "Mic", cleaning_event_col = "CleaningEvent")
print(ewma_plot)

Poisson Fixed Effect Model Summary

Description

Fits a simple Poisson model to the data and returns a data frame containing the model's term, estimate, standard error, z value, and p-value, formatted to a fixed number of decimal places.

Usage

cv19_poisson_simple(data, residue_col)

Arguments

Value

A data frame with the formatted summary of the Poisson regression model.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_Mic' is a data frame and 'Mic' is the residue column of interest.
cv19_poisson_simple(data = Eq_Mic, residue_col = "Mic")

Poisson Fixed Effect Model

Description

Fits a fixed effects Poisson model and returns a data frame with the summary. If the p-value is significant, then the corresponding cleaning event is significantly different from other cleaning events. For a stable cleaning process, we wish all p-values are not significant.

Usage

cv20_poisson_fixed(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame output with the fixed effect summary.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

fixed_effect_summary <- cv20_poisson_fixed(data = Eq_Mic, residue_col = "Mic", 
cleaning_event_col = "CleaningEvent")

Poisson Mixed Effect Model Summary

Description

Fits a mixed-effects Poisson model to the data and returns a data frame containing the fixed effect part estimates, standard errors, z-values, and p-values.

Usage

cv21_poisson_mixed(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame with the fixed effect summary of the mixed-effects Poisson regression model.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming 'Eq_Mic' is a data frame, 'Mic' is the residue column of interest,
# and 'CleaningEvent' is the column for random effects grouping.
mixed_effect_summary <- cv21_poisson_mixed(data = Eq_Mic, residue_col = "Mic", 
cleaning_event_col = "CleaningEvent")
print(mixed_effect_summary)

Extract Variance of Random Effects

Description

This function fits a Poisson mixed-effects model with a specified random effect and extracts the variances and standard deviations of the random effects.

Usage

cv22_var_random_effect(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame with the variances and standard deviations of the random effects.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

ra_table <- cv22_var_random_effect(data=Eq_Mic, residue_col="Mic", 
cleaning_event_col="CleaningEvent")

Extract Random Effect Coefficients

Description

This function fits a Poisson mixed-effects model with a specified random effect and extracts the random effect coefficients and their standard deviations.

Usage

cv23_random_effect_coef(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame with the random effect coefficients and standard deviations.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

re_coefs <- cv23_random_effect_coef(data=Eq_Mic, residue_col="Mic", 
cleaning_event_col="CleaningEvent")

Variance Component Analysis for Microbial Counts

Description

This function performs a variance component analysis using a mixed-effects model with a Poisson distribution to estimate within-group and between-group variance for microbial counts data. Assumes data is grouped by cleaning events and evaluates the residue or microbial counts within these groups.

Usage

cv24_vca_mic(data, residue_col, cleaning_event_col)

Arguments

Value

A data frame summarizing the variance components, including within-group variance, between-group variance, and total variance, along with their percentages and standard deviations.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming `Eq_Mic` is your dataframe, `Mic` is the microbial counts column, 
# and `CleaningEvent` is the cleaning event column:
cv24_vca_mic(Eq_Mic, "Mic", "CleaningEvent")

Binomial Process Performance Calculation

Description

Performs a process performance calculation using binomial distribution. This includes a bootstrap procedure to estimate the confidence interval of the Process Performance Index (Ppu).

Usage

cv25_qbinom_ppu(data, residue_col, cleaning_event_col, usl_col)

Arguments

Value

A data frame with the calculated Ppu and its 95 along with the method used ("Q-Binomial").

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

# Assuming `data` is the dataframe with columns "Residue", "CleaningEvent", and "USL":
cv25_qbinom_ppu(Eq_Mic, "Mic", "CleaningEvent", "USL")

Calculate Process Performance Index using Poisson Distribution

Description

This function calculates the Process Performance Index (Ppu) for data assumed to follow a Poisson distribution. It includes a bootstrap method for estimating the confidence interval of the Ppu.

Usage

cv26_qpoisson_ppu(data, residue_col, cleaning_event_col, usl_col)

Arguments

Value

A data frame with columns Method, Ppu, CI_Lower, and CI_Upper.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv26_qpoisson_ppu(Eq_Mic, "Mic", "CleaningEvent", "USL")

Process Performance Calculation using Anscombe's Transformation

Description

Calculates the Process Performance Index (Ppu) using Anscombe's transformation. This function also performs a bootstrap to estimate the confidence interval of Ppu.

Usage

cv27_anscombe_ppu(data, residue_col, cleaning_event_col, usl_col)

Arguments

Value

A data frame with columns for the Method, Ppu, CI_Lower, and CI_Upper.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv27_anscombe_ppu(Eq_Mic, "Mic", "CleaningEvent", "USL")

Calculate Ppu using Freeman's Transformation

Description

This function calculates the Process Performance Index (Ppu) using Freeman's transformation, including a bootstrap method to estimate the confidence interval of Ppu.

Usage

cv28_freeman_ppu(data, residue_col, cleaning_event_col, usl_col)

Arguments

Value

A data frame with columns for the Method, Ppu, CI_Lower, and CI_Upper.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

cv28_freeman_ppu(Eq_Mic, "Mic", "CleaningEvent", "USL")

Calculate Mic Ppu with Five Methods

Description

This function calculates the process performance index (Ppu) for Mic using five different methods, including Q-Binomial, Q-Poisson, Anscombe, Freeman, and KDE. It returns a dataframe with the Ppu values, lower and upper confidence intervals for each method, and appends a row for the method with the minimum Ppu value.

Usage

cv29_mic_ppu(data, residue_col, cleaning_event_col, usl_col)

Arguments

Value

A dataframe with the Ppu calculations for each method and the minimum Ppu method.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

  MicPPU <- cv29_mic_ppu(Eq_Mic, "Mic", "CleaningEvent", "USL")

Calculate DAR, CAR, and Mic Ppu Values and Identify the Overall Minimum

Description

This function calculates Ppu values for DAR, CAR, and Mic using the KDE method provided by the 'cv12_kde_ppu' function. It then uses the 'cv29_mic_ppu' function to calculate combined Ppu for Mic and extract the method with the minimum Ppu value. The function assumes the availability of the datasets 'Eq_DAR', 'Eq_CAR', and 'Eq_Mic' that conform to expected column naming conventions and data structures. It is reliant on the results of the 'cv12_kde_ppu' and 'cv29_mic_ppu' functions being consistent and correctly formatted.

Usage

cv30_dar_car_mic_ppu(
  dar_data,
  dar_residue_col,
  dar_cleaning_event_col,
  dar_usl_col,
  car_data,
  car_residue_col,
  car_cleaning_event_col,
  car_usl_col,
  mic_data,
  mic_residue_col,
  mic_cleaning_event_col,
  mic_usl_col
)

Arguments

Value

A dataframe with Ppu values for DAR, CAR, and Mic, along with the Overall Minimum Ppu.

Author(s)

Chan, Mohamed, Lou, Wendy, Yang, Xiande [xiande.yang at gmail.com]

Examples

Eq_DAR <- cv03_usl_unification(data = Eq_DAR, cleaning_event_col = "CleaningEvent", 
 residue_col = "DAR", usl_col = "USL")
 Eq_CAR <- cv03_usl_unification(data = Eq_CAR, cleaning_event_col = "CleaningEvent", 
 residue_col = "CAR", usl_col = "USL")
  df1 <- cv30_dar_car_mic_ppu(
    dar_data = Eq_DAR, 
    dar_residue_col = "DAR_Pct", 
    dar_cleaning_event_col = "CleaningEvent", 
    dar_usl_col = "USL_Pct",
    car_data = Eq_CAR, 
    car_residue_col = "CAR_Pct", 
    car_cleaning_event_col = "CleaningEvent", 
    car_usl_col = "USL_Pct",
    mic_data = Eq_Mic, 
    mic_residue_col = "Mic", 
    mic_cleaning_event_col = "CleaningEvent", 
    mic_usl_col = "USL")