Wilcoxon Signed Rank Test

hanaml.Wilcoxon is a R wrapper for SAP HANA PAL Wilcox Signed Rank Test.

hanaml.Wilcoxon(
  data,
  col = NULL,
  mu = NULL,
  test.type = NULL,
  correction = NULL
)

Arguments

data	DataFrame DataFrame containting the data.
col	character, optional Name of the data column that needs to be tested. If not given, the input dataframe must only have one column. Defaults to NULL.
mu	double, optional Specifies the hypothetical value of median. Defalut to 0.
test.type	{'two.sides', 'less','greater'},optional Specifies the type of the hypothesis for the test. The types are given as follows: 'two.sides':The true median is different from mu. 'less':The true median is less than mu. 'greater':The true median is greater than mu. Defaults to 'two.sides'.
correction	logical, optional Specifies whether or not to include the continuity correction for the p value calculation. Defaults to TRUE.

Value

• DataFrame
  Statistical information of the test,e.g. p-value

  • STAT_NAME : Statistics name.

  • STAT_VALUE : Statistics value.

Details

Wilcoxon signed rank test is a non-parametric procedure which tests the sample median against a hypothetical median. The Wilcoxon signed rank test should be used if the sample data come from a symmetrical distribution, otherwise it is recommended to use the one-sample-median test.

Examples

Input DataFrame data of one-sample example:

> data$Head(5)$Collect()
   X
1 85
2 65
3 20
4 56
5 30

Call the function:

> result <- hanaml.Wilcoxon(data = data, test.type = "two.sides",
                          mu = 40, correction = TRUE )

Results:

> result$Collect()
  STAT_NAME   STAT_VALUE
1 statistic 158.50000000
2   p-value   0.01122824

Input DataFrame data of paired two-sample example:

> data$Head(5)$Collect()
  X1 X2
1 85 60
2 65 68
3 20 20
4 56 89
5 30  7

Call the function:

> result <- hanaml.Wilcoxon(data = data, test.type = 'two.sides' ,
                            correction = TRUE)

Results:

> result$Collect()
  STAT_NAME  STAT_VALUE
1 statistic 118.5000000
2   p-value   0.3546207