Pandas types

Pandas DataFrame and Series are commonly used data structures. The Pandas types modules provide type converters for DataFrame’s and Series within PyXLL.

It provides the following bi-directional type converters:

• dataframe: converter for the pandas.DataFrame objects with support for multi-index or columns
• series: converter for the pandas.Series objects mapped to a column vector in Excel
• series_t: converte for the pandas.Series objects mapped to a row vector in Excel

How to use it?

Once you have loaded the pyxll_utils.pandastypes module, you can use the custom type converters. The following example shows how to use DataFrames as input and output:

    from pyxll import xl_func
    import pandas as pa

    @xl_func("int rows, int cols, float value: dataframe")
    def make_empty_dataframe(rows, cols, value):
            # create an empty dataframe
            df = pa.DataFrame({chr(c + ord('A')) : value for c in range(cols)}, index=range(rows))

            # return it. The custom type will convert this to a 2d array that
            # excel will understand when this function is called as an array
            # function.
            return df

@xl_func("dataframe df, string col: float")
def sum_column(df, col):
    return df[col].sum()