bnrow

Counts the number of rows using data.table::fread() and the 'wc' Unix command.
This allows analyzing big files that would not fit in memory (and cause an error
such as 'cannot allocate vector of size').

A simple set of wrapper functions for data.table::fread() that allows subsetting or
filtering rows and selecting columns of table-formatted files too large for the available RAM.
'b stands for 'big files'.
bread makes heavy use of Unix commands like 'grep', 'sed', 'wc', 'awk' and 'cut'. They are available
by default in all Unix environments.
For Windows, you need to install those commands externally in order to simulate a
Unix environment and make sure that the executables are in the Windows PATH variable.
To my knowledge, the simplest ways are to install 'RTools', 'Git' or 'Cygwin'. If they have been
correctly installed (with the expected registry entries), they should be detected on loading
the package and the correct directories will be added automatically to the PATH.

Vincent Guegan

bread

Analyze Big Files Without Loading Them in Memory

bnrow function

<dl><dt>file</dt>
<dd>String. Name or full path to a file compatible with data.table::fread()</dd></dl>

Arguments

Count the number of rows of a big file without loading it in memory — bnrow

<dl>

<dt>file</dt>
<dd>String. Name or full path to a file compatible with data.table::fread()</dd>

</dl>

bnrow: Count the number of rows of a big file without loading it in memory

Description

Usage

Value

Arguments

Examples