Cleaning, splicing, and organizing arrays

You can often think of all the data in a project you are dealing with as a giant table. For instance the table will be indexed by things like configuration number, ensemble stream/series, quark mass, and so on. Sometimes it is important to find data that lies at the intersection of some sets, or to look only at data whose bare coupling \(\beta\) fall within some prescribed range, and so on. The methods in

latqcdtools.base.cleanData

try to help with this.

These manipulations are already possible within numpy, but they can be difficult to read. (At least they are difficult for David to read.) Hence this module contains wrappers for such manipulations. They include

• deleteRow: Delete a row of a 2-d array.

• deleteCol: Delete a column of a 2-d array.

• clipRange: Pick a column col of a data table data. Throw out all data whose corresponding entries in col fall outside the range [minVal,maxVal].

• intersectAtCol: Return only those rows of table1 and table2 that have identical elements in column col.

• spliceAtCol: Assuming two tables table1 and table2 have common values in column col, create a new table, where table1 has corresponding entries less than atVal in col, and table2 has corresponding entries greater than `atVal.

• restrictAtCol: Return only those rows of a table where col has exactly the value atVal.

• excludeAtCol: Return everything except those rows of a table where col has exactly the value atVal.