Use Sphinx's :manpage: construct for python(1).

[hooke.git] / doc / tutorial.txt

********
Tutorial
********

`A short video showing Hooke in action`_! (courtesy of Fabrizio
Benedetti, EPFL, Lausanne)

.. _A short video showing Hooke in action:
  https://documents.epfl.ch/users/f/fb/fbenedet/www/hooke_short_demostration.ogv

.. toctree::
   :maxdepth: 2

   gui

Introduction
============

This tutorial will focus on the command-line interface as the most
powerful, and leave the :doc:`gui` interface to another document.

.. _command-line: `Command-line interface`_

Installation
============

See :doc:`install` for details on downloading and installing Hooke.


Command-line interface
======================

Running the hooke shell
-----------------------

Hooke has a set of commands that depend on the loaded
:class:`hooke.plugin.Plugin`\s.  To access these commands, you'll need
to run the Hooke shell.::

    $ hooke

If you are running hooke from the source directory (see
:doc:`install`), the equivalent command is::

    $ python bin/hooke

You may need to give the full path for Python on Windows systems, and
also check that the current working directory (`.`) is in your
`PYTHONPATH`.  See :manpage:`python(1)` for details.

As Hooke launches, you should see something like the following in your
terminal::

    Hooke version 0.8.0 Seinei

    COPYRIGHT
    ----
    hooke>

The final line, ``hooke>``, is the Hooke prompt.  It allows you to
enter commands to interact with the interpreter.

Help
----

All commands have help text explaining their purpose and usage.  The
text is stored in the code itself, and therefore more likely to be up
to date than this tutorial.  You can get a list of commands and topics
with::

    hooke> help

Or see specific help on ``TOPIC`` with::

    hooke> help TOPIC

for example::

    hooke> help load_playlist

will give help on the ``load_playlist`` command.

Leaving Hooke
-------------

When you're done with an interactive Hooke session, you can close the
session with ``exit`` or its aliases ``quit`` and ``EOF`` (``EOF`` is
the end of the stdin stream, which is Ctrl-d in many shells).

Creating a playlist
-------------------

To start analyzing your curves, you first have to build a playlist. The
playlist is just an index of the force curve files you want to
analyze. Imagine it as a music playlist (that’s why it is called a
playlist), but with data files instead of audio files.

Suppose you have 100 PicoForce curve files in your curves directory,
starting from :file:`mycurve.000` and ending in :file:`mycurve.100`
and you want to analyze them all.

You then can ``cd`` (change directory) to the directory::

    hooke> cd --path c:\curves

Type ``pwd`` (print working directory) to check the directory is correct.::

    hooke> pwd
    c:\curves

You can list the files in the directory using ``ls`` or ``dir``
(they’re synonyms).::

    hooke> ls
    mycurve.000
    mycurve.001
    ...

Now you are ready to generate the playlist. The command to use is
``glob_curves_to_playlist``.::

    hooke> glob_curves_to_playlist mycurve.*

You can also be more specific with wildcards.  For example

    hooke> glob_curve_to_playlist mycurve.05*

will take only curves from :file:`mycurve.050` to :file:`mycurve.059`.

Note that by using ``glob_curve_to_playlist`` you just generate the
playlist in the local session. To save your playlist to a file for
future reuse, type::

    hooke> save_playlist --output mylist

In this example, the list will be saved in the file
:file:`mylist.hkp`.  Hooke will add the extension ``.hkp`` (Hooke
playlist) to the playlist if you forget to.  The ``.hkp`` file is an
XML file you can read and edit with any text editor (i.e. Wordpad), if
needed.  If you want to load it, simply issue ``load_playlist
mylist.hkp`` or ``load_playlist mylist``, Hooke will add ``.hkp`` if
necessary.

If, generating the playlist, you are including by chance a non-force
curve file that Hooke cannot open, Hooke will print an error and
continue on.

Navigating the playlist
-----------------------

Now you can navigate through your playlist using the commands
``next_curve`` and ``previous_curve``. You don’t need to type
``next_curve`` every time to run along a list of curves.  If you press
Return to an empty prompt, Hooke will repeat the last command you
issued explicitly.  You can also navigate through the command history
by using the up and down arrows, or auto-complete partial commands
with TAB.  From the last curve of your playlist, ``next_curve`` will
wrap around to the first curve.  Analogously, issuing
``previous_curve`` at the first curve will jump to the last.

You can also jump to a given curve::

    hooke> jump_to_curve 14

will jump to the 14th curve in the playlist.  TODO: jump_to_curve
<PATH>, where the path can be either an absolute path or a path
relative to the directory holding the playlist file.

Taking notes
------------

You can take notes about the curves you are looking at.  Just type
``add_note`` followed by the text you want to append to that curve.
Hooke will save the text in your current playlist and in an external
log file (TODO: no external file yet.  Is this important?).  The
output will look like this::

    Notes taken at Sun Sep 17 20:42:07 2006
    /home/cyclopia/work/tris/20060620a.041 | This is a note
    /home/cyclopia/work/tris/20060620a.207 | This is another note
    /home/cyclopia/work/tris/20060620a.286 | This is a third one

The log file name can be configured (:doc:`config`), but it defaults
to :file:`hooke.log`.

Usually curves you annotated are useful later.  You can create a
playlist for only annotated curves with

    hooke> note_filter_playlist --name c:\curves\nice.hkp

will create sub-playlist :file:`c:\curves\nice.hkp`.  Make sure that
the target directory (here :file:`c:\curves\`) already exists before
doing that.

If you change your mind about a note, you can remove it with
``clear_note``.

Exporting curves
----------------

You can export Hooke curves as images and as text columns. To export
as images or text, use the ``export_block`` command.  Supported
formats are PNG (Portable Network Graphic, raster) and EPS
(Encapsulated Postscript, vector).  The export format is determined by
the filename extension, so ``export_block foo.png``, ``export_block
foo.eps``, and ``export_block foo.txt`` will save PNG, EPS, and
TAB-delimited text files respectively.

TODO: multiple cycles?  Solution: blank lines for "breaks", add option
to extract specific sections using Python's slice notation.

If you don't want the entire block, try the ``cut`` command.

Interacting with the plot
-------------------------

(no plots in command line mode...)

Measuring distances and forces
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

You can easily zoom in the plot by dragging a rectangle on it with the
left mouse button.  To zoom out, click the right mouse
button. Sometimes by zooming in and out too much, you can lose the
picture (this is probably a small bug in Matplotlib).  Just type
``plot`` at the command line and the curve will be refreshed.

You can measure distances and forces directly in the plot. Just issue
the command ``distance``.  You will be asked to click two points.
When you click a point, a blue dot should appear.  When you click the
second point, the distances (in nanometers and piconewtons) will
appear on the command line.  You can use ``delta`` if you prefer,
which gives meaningful values for every kind of graph (not only force
curves). If you want to know the coordinates of a single point, use
``point``.

Hooke automatically adjusts the position of the clicked point to the
nearest point in the graph, so you will be always measuring distances
and forces between points in the graph.

The commands ``force`` and ``distance`` are present in the
``generalvclamp`` plugin.

Worm like chain and freely jointed chain fitting
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

You can measure by hand the parameters relative to a force peak using
a worm-like chain fitting with the ``fit`` command.  The command by
default automatically finds the contact point, asks for two points
delimiting the portion to fit, and performs a two-variable fit, with
contour length, persistence length, and their relative errors as
output.  If desired, one can use the ``noauto`` option to manually
click the contact point, and/or the ``pl=NUMBER`` options to impose a
specific persistence or Kuhn length (in nanometers). You can choose
which model to use with ``set fit_function wlc`` or ``set fit_function
fjc``.  See the help of the ``fit`` command from the Hooke command
line for details.

Multiple curve fitting and measuring
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

You can cycle through all your current playlist obtaining WLC fit, FJC
fit, rupture force and slope (loading rate) information from each
curve using the ``multifit`` command.  The collected data can be saved
in a text file for further analysis in your favourite spreadsheet or
statistical program.  If you want to check your parameters on the
current curve before fitting all the files in your playlist, use
``multifit justone``.  See the ``multifit`` help for more options.

Fast curve reviewing and saving
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

When automatic routines are not good enough to filter your data, use
``review`` command to cycle through your playlist presenting ten
curves in the same graph.  You can then enter the numbers of the
interesting curves and automatically save a copy of them into another
directory.

Configuring Hooke
-----------------

You can set environment variables to influence the behaviour of
Hooke. The command to use is ``set_config``.

You can alter permanently the behaviour of Hooke by setting these
variables in a Hooke configuration file.  See :doc:`config` for
details.