public static TreeScore NumericHypothesisTest(DAG dag, AST.Range rangeNode, AST.Address functionNode, FunctionOutput <string>[] boots, string initial_output, bool weighted, double significance)
{
// this function's input cells
var input_cells = rangeNode.Addresses();
var inputs_sz = input_cells.Count();
// scores
var input_exclusion_scores = new TreeScore();
// convert to numeric
var numeric_boots = ConvertToNumericOutput(boots);
// sort
var sorted_num_boots = SortBootstraps(numeric_boots);
// for each excluded index, test whether the original input
// falls outside our bootstrap confidence bounds
for (int i = 0; i < inputs_sz; i++)
{
// default weight
int weight = 1;
// add weight to score if test fails
AST.Address xtree = input_cells[i];
if (weighted)
{
// the weight of the function value of interest
weight = dag.getWeight(functionNode);
}
double outlieriness = RejectNullHypothesis(sorted_num_boots, initial_output, i, significance);
if (outlieriness != 0.0)
{
// get the xth indexed input in input_rng i
if (input_exclusion_scores.ContainsKey(xtree))
{
input_exclusion_scores[xtree] += (int)(weight * outlieriness);
}
else
{
input_exclusion_scores.Add(xtree, (int)(weight * outlieriness));
}
}
else
{
// we need to at least add the value to the tree
if (!input_exclusion_scores.ContainsKey(xtree))
{
input_exclusion_scores.Add(xtree, 0);
}
}
}
return(input_exclusion_scores);
}
public static TreeScore StringHypothesisTest(DAG dag, AST.Range rangeNode, AST.Address functionNode, FunctionOutput <string>[] boots, string initial_output, bool weighted, double significance)
{
// this function's input cells
var input_cells = rangeNode.Addresses();
// scores
var iexc_scores = new TreeScore();
var inputs_sz = input_cells.Count();
// exclude each index, in turn
for (int i = 0; i < inputs_sz; i++)
{
// default weight
int weight = 1;
// add weight to score if test fails
AST.Address xtree = input_cells[i];
if (weighted)
{
// the weight of the function value of interest
weight = dag.getWeight(functionNode);
}
if (RejectNullHypothesis(boots, initial_output, i, significance))
{
if (iexc_scores.ContainsKey(xtree))
{
iexc_scores[xtree] += weight;
}
else
{
iexc_scores.Add(xtree, weight);
}
}
else
{
// we need to at least add the value to the tree
if (!iexc_scores.ContainsKey(xtree))
{
iexc_scores.Add(xtree, 0);
}
}
}
return(iexc_scores);
}
public static TreeScore StringHypothesisTest(DAG dag, AST.Range rangeNode, AST.Address functionNode, FunctionOutput<string>[] boots, string initial_output, bool weighted, double significance)
{
// this function's input cells
var input_cells = rangeNode.Addresses();
// scores
var iexc_scores = new TreeScore();
var inputs_sz = input_cells.Count();
// exclude each index, in turn
for (int i = 0; i < inputs_sz; i++)
{
// default weight
int weight = 1;
// add weight to score if test fails
AST.Address xtree = input_cells[i];
if (weighted)
{
// the weight of the function value of interest
weight = dag.getWeight(functionNode);
}
if (RejectNullHypothesis(boots, initial_output, i, significance))
{
if (iexc_scores.ContainsKey(xtree))
{
iexc_scores[xtree] += weight;
}
else
{
iexc_scores.Add(xtree, weight);
}
}
else
{
// we need to at least add the value to the tree
if (!iexc_scores.ContainsKey(xtree))
{
iexc_scores.Add(xtree, 0);
}
}
}
return iexc_scores;
}
public static TreeScore NumericHypothesisTest(DAG dag, AST.Range rangeNode, AST.Address functionNode, FunctionOutput<string>[] boots, string initial_output, bool weighted, double significance)
{
// this function's input cells
var input_cells = rangeNode.Addresses();
var inputs_sz = input_cells.Count();
// scores
var input_exclusion_scores = new TreeScore();
// convert to numeric
var numeric_boots = ConvertToNumericOutput(boots);
// sort
var sorted_num_boots = SortBootstraps(numeric_boots);
// for each excluded index, test whether the original input
// falls outside our bootstrap confidence bounds
for (int i = 0; i < inputs_sz; i++)
{
// default weight
int weight = 1;
// add weight to score if test fails
AST.Address xtree = input_cells[i];
if (weighted)
{
// the weight of the function value of interest
weight = dag.getWeight(functionNode);
}
double outlieriness = RejectNullHypothesis(sorted_num_boots, initial_output, i, significance);
if (outlieriness != 0.0)
{
// get the xth indexed input in input_rng i
if (input_exclusion_scores.ContainsKey(xtree))
{
input_exclusion_scores[xtree] += (int)(weight * outlieriness);
}
else
{
input_exclusion_scores.Add(xtree, (int)(weight * outlieriness));
}
}
else
{
// we need to at least add the value to the tree
if (!input_exclusion_scores.ContainsKey(xtree))
{
input_exclusion_scores.Add(xtree, 0);
}
}
}
return input_exclusion_scores;
}