robust.categorical.igate {igate}

Robust igate for categorical target variables

Description

This function performs a robust an initial Guided Analysis for parameter testing and controlband extraction (iGATE) for a categorical target variable by repeatedly running categorical.igate and only returning those parameters that are selected more often than a certain threshold.

Usage

robust.categorical.igate(df, versus = 8, target, best.cat, worst.cat,
  test = "w", ssv = NULL, outlier_removal_ssv = TRUE,
  iterations = 50, threshold = 0.5)

Arguments

df

Data frame to be analysed.

versus

How many Best of the Best and Worst of the Worst do we collect? By default, we will collect 8 of each.

target

Target variable to be analysed. Must be categorical. Use igate for continuous target.

best.cat

The best category. The versus BOB will be selected randomly from this category.

worst.cat

The worst category. The versus WOW will be selected randomly from this category.

test

Statistical hypothesis test to be used to determine influential process parameters. Choose between Wilcoxon Rank test ("w", default) and Student's t-test ("t").

ssv

A vector of suspected sources of variation. These are the variables in df which we believe might have an influence on the target variable and will be tested. If no list of ssv is provided, the test will be performed on all numeric variables.

outlier_removal_ssv

Logical. Should outlier removal be performed for each ssv (default: TRUE)?

iterations

Integer. How often should categorical.igate be performed? A message about how many iterations have been perfermed so far will be printed to the console every 0.1*iterations iterations.

threshold

Between 0 and 1. Only parameters that are selected at least floor(iterations*threshold) times are returned.

Details

We collect the Best of the Best and the Worst of the Worst dynamically dependent on the current ssv. That means, for each ssv we first remove all the observations with missing values for that ssv from df. Then, based on the remaining observations, we randomly select versus observations from the the best category (“Best of the Best”, short BOB) and versus observations from the worst category (“Worst of the Worst”, short WOW). By default, we select 8 of each. Since this selection happens randomly, it is recommended to use robust.categorical.igate over categorical.igate. After the selection we compare BOB and WOW using the the counting method and the specified hypothesis test. If the distributions of the ssv in BOB and WOW are significantly different, the current ssv has been identified as influential to the target variable. An ssv is considered influential, if the test returns a count larger/ equal to 6 and/ or a p-value of less than 0.05. For the next ssv we again start with the entire dataset df, remove all the observations with missing values for that new ssv and then select our new BOB and WOW. In particular, for each ssv we might select different observations. This dynamic selection is necessary, because in case of an incomplete data set, if we select the same BOB and WOW for all the ssv, we might end up with many missing values for particular ssv. In that case the hypothesis test loses statistical power, because it is used on a smaller sample or worse, might fail altogether if the sample size gets too small.

For those ssv determined to be significant, control bands are extracted. The rationale is: If the value for an ssv is in the interval [good_lower_bound,good_upper_bound] the target is likely to be good. If it is in the interval [bad_lower_bound,bad_upper_bound], the target is likely to be bad.

Value

A data frame with the summary statistics for those parameters that were selected at least floor(iterations*threshold) times:

Causes	Those ssv that have been found to be influential to the target variable.
median_count	The median value returned by the counting method for this parameter.
median_p_value	The median p-value of the hypothesis test performed, i.e. either of the Wilcoxon rank test (in case test = "w") or the t-test (if test = "t").
median_good_lower_bound	The median lower bound for this Cause for good quality.
median_good_upper_bound	The median upper bound for this Cause for good quality.
median_bad_lower_bound	The median lower bound for this Cause for bad quality.
median_bad_upper_bound	The median upper bound for this Cause for bad quality.

Examples

robust.categorical.igate(mtcars, target = "cyl",
best.cat = "8", worst.cat = "4", iterations = 50, threshold = 0.5)