Score

Public Methods

• Score() - Default Constructor.
• Score(Score) - Copy Constructor.
• Score(string) - Loads the Score from a file.
• Copy(Score) - Copy a Score.
• SetDataPoint(int, int, double) - Sets a data point.
• GetDataPoint(int, int) - Gets a data point.
• Load(string) - Loads the object from the file given in the file path.
• Save(string) - Writes the data to the file given in the file path.
• AppendSave(string) - Appends the Score to the file given in the file path.
• AppendSave(string, char) - Appends the Score to the file given in the file path that uses the given delimiter.
• Filter(string) - Filters out rows based on a valid function.
• SignChange(string, arrayList) - Check to see that "lead" represents a valid column name. For each row, get the sign of the lead column, and set the specified columns to the same sign. Will modify the data!
• Standardize(ArrayList) - For each column, create a new column "columnName_std". For each of the new columns, get the standard deviation and the mean. For each row in the "_std" columns, modify the value by (the mean and standard deviation will be column specific). Will modify the data!
• GenerateRankings(CriteriaSet) - Has the columns to rank and asc or desc as their operator. Defaults to no cross-validation.
• GenerateRankings(CriteriaSet, int) - Has the columns to ranks and asc or desc as their operator. Takes an integer for cross-validation.
• GenerateRankings(ArrayList, ArrayList) - Ranks scores ascending or descending. Defaults to no cross-validation.
• GenerateRankings(ArrayList, ArrayList, int) - For each column that is listed, sort the score ascending or descending according to the correspoinding specification. Create a new column "columnName_rank" for each score that is given. Assign a numeric value 1 - n for each row id. If there is a tie, the row ids get the same rank until the tie is broken. The score should resume as though there is no interruption is ranks. Ex: 1, 2, 3, 3, 3, 6, 7. After individual colums have been ranked, create a new column "rank_sum" The rank sum is the sum of the ranks. Sort rank_sum and create a new column "rank" which is the rank of rank_sum. Ties in rank sum are broken by the first column in the scoreList. Ties that remain are broken by the row_id. Takes an integer for cross-validation.

Private Methods

• RankByColumn(int, bool, int) - Generates the ranks for a column in another column. There is no tie-breaking.

Virtual Methods

• GenerateScores(DataSet, DataGrouping) - Place holder for child class implementations. Defaults to no cross-validation.
• GenerateScores(DataSet, DataGrouping, int) - Place holder for child class implementation. Takes an integer for cross-validation.

Exceptions

• Standerdize/signchange tries to modify a column that does not exist.
• The grouping is not correctly set up. E.g. Missing group 1 or 2. Grouping other the 1, 2 or 0.
• A file is not formatted correctly.
• A file to read does not exist.

Remarks

• All columns must be group 1, 2, or 0. The data grouping must have groups 1, 2. Group 0 is optional.
• Missing values: stored as Double.NaN, written as ".".
• Delimiters: tab(default) or space.