source: branches/numpy/anuga/utilities/csv_tools.tex @ 9053

\documentclass{manual}

\title{Module csv\_tools}

\author{ANUGA Developer}

\begin{document}
\maketitle

\chapter{Module \code{csv_tools} functions}

This document describes the functions within the \code{csv_tools.py} module.
The \LaTeX is here as a placeholder while it is decided where this sort of things goes.

\section{\code{merge_csv_key_values()}}

\begin{methoddesc}{merge_csv_key_values}{file_title_list,
                                         output_file,
                                         key_col='hours',
                                         data_col='stage'}
Module: \module{csv_tools}

Merge one or more CSV files into a single output CSV file.  The output file contains a \emph{key}
column that is common between all input files and one \emph{data} column from each input file.

\code{file_title_list} is a list of 2-tuples, each containing the path to an input file and a
new column header string for the \emph{data} column from the file.

\code{output_file} is the path to the output file.

\code{key_col} is the column header string that identifies the \emph{key} column in each
input file.  If not provided, the \emph{key} column has the string 'hours'.

\code{data_col} is the column header string that identifies the \emph{data} column in each
input file.  If not provided, the \emph{data} column has the header string 'stage'.

As an example, suppose we have two CSV files \code{alpha.csv}:

\begin{table}[htp]
  \begin{center}
    \begin{tabular}{|cccc|}
      \hline
      time & hours & stage & depth \\
      \hline
      3600 & 1.00 & 100.3 & 10.2 \\
      3636 & 1.01 & 100.3 & 10.0 \\
      3672 & 1.02 & 100.3 &  9.7 \\
      3708 & 1.03 & 100.3 &  8.9 \\
      3744 & 1.04 & 100.3 &  7.1 \\
      \hline
    \end{tabular}
  \end{center}
\end{table}

and \code{beta.csv}:

\begin{table}[htp]
  \begin{center}
    \begin{tabular}{|cccc|}
      \hline
      time & hours & stage & depth \\
      \hline
      3600 & 1.00 & 100.3 & 11.3 \\
      3636 & 1.01 & 100.3 & 10.5 \\
      3672 & 1.02 & 100.3 & 10.0 \\
      3708 & 1.03 & 100.3 &  9.7 \\
      3744 & 1.04 & 100.3 &  8.2 \\
      \hline
    \end{tabular}
  \end{center}
\end{table}

and we wish to merge these two files, using the \code{hours} column as the \emph{key} and
the \code{depth} column as \emph{data} in both input files.

We would do this with the code fragment:

\begin{verbatim}
title_list = [('alpha.csv', 'alpha'), ('beta.csv', 'beta')}
output_file = 'gamma.csv'
merge_csv_key_values(title_list, output_file, key_col='hours', data_col='depth')
\end{verbatim}

The output file \code{gamma.csv} would contain:

\begin{table}[htp]
  \begin{center}
    \begin{tabular}{|ccc|}
      \hline
      hours & alpha & beta \\
      \hline
      1.00 & 10.2 & 11.3 \\
      1.01 & 10.0 & 10.5 \\
      1.02 &  9.7 & 10.0 \\
      1.03 &  8.9 &  9.7 \\
      1.04 &  7.1 &  8.2 \\
      \hline
    \end{tabular}
  \end{center}
\end{table}

The function looks for the \emph{key} column in all input files and ensures it is the same in all input files.

The \emph{data} column must exist in all input files, and the column for each file is copied to the 
output file.  The column header string is changed to be the string entered against each input filename 
in the \code{title_list} 2-tuple for the input file.
\end{methoddesc}

\end{document}