ipython_ferretmagic

IPython notebook extension for ferret (a great software for Data Visualisation and Analysis from NOAA/PMEL).

Browse the demo notebooks from http://nbviewer.ipython.org/github/PBrockmann/ipython_ferretmagic/tree/master/notebooks

Combine ferret and ipython widgets and create interactive notebooks !

Test it with binder.

Browse a basic ferret tutorial using ferretmagic from: https://github.com/PBrockmann/C-CASCADES_notebooks/blob/master/ferretBasics.ipynb

Latest news

2019/11/13

• Tested with python 3.7, pyferret 7.5.0
• Notebooks run with mybinder.org with 2 files (environment.yml and postBuild)

2018/12/10

• Tested with pyferret==7.44

2018/10/01

• Compatibility with python 3 with print command and base64 encoding
• Tested with pyferret==7.43, notebook==5.7.0, jupyter==1.0.0, ipython==5.8.0, ipywidgets==7.4.2
• Update pip package (now last version is 20181001 https://pypi.org/project/ferretmagic/)

2018/01/18

• a Docker image is available from https://hub.docker.com/r/pbrockmann/pyferret-7.3/
• pyferret 7.3 and ferretmagic extension can now be fully tested from binder (http://mybinder.readthedocs.io/en/latest) !
• Tested with pyferret==7.3, jupyter==4.4.0, ipython==5.5.0, ipywidgets==7.0.5

2017/03/17

• Tested with pyferret==7.1, jupyter==4.2.0, ipywidgets==6.0.0

2016/10/28

• Tested with pyferret==7.0, jupyter==4.2.0 and ipython==5.1.0
• Installation from pip

Installation

Install PyFerret following instructions from http://ferret.pmel.noaa.gov/Ferret/documentation/pyferret It will set all directories needed to run pyferret (font, ppl, jnls, datasets, ...).

You can also install PyFerret as a conda package from https://anaconda.org/conda-forge/pyferret (no datasets distributed).

Then install the ferret magic extension (ferretmagic.py) by typing pip install ferretmagic

Usage

In a Jupyter notebook with kernel python (previously named ipython), load the magics:

In [2]: %load_ext ferretmagic

Line magics

The %ferret_run magic enables one-line execution of ferret command in the IPython interpreter or notebook:

    In [3]: for i in [100,500,1000]:
      ....: 	%ferret_run -s 400,400 'plot sin(i[i=1:%(i)s]*0.1)' % locals()

Cell magics

Multi-line input can be entered with the %%ferret cell magic:

In [4]: %%ferret
  ....: use levitus_climatology
  ....: shade temp[k=1]			! comments

• Control size of plot with --size, -s option.
• Create a local pdf at the same time with --pdf, -p option.

In [5]: %%ferret -s 400,300 -pdf myfig.pdf
  ....: shade temp[k=1]			

Passing variables between Python and ferret

• Variables can be pushed from IPython into ferret with %ferret_putdata:

In [18]: %%ferret
   ....: use levitus_climatology
In [19]: %ferret_getdata tempdict = temp
   ....: Message: tempdict is now available in python as a dictionary containing the variable's metadata and data array.
In [20]: print tempdict.keys()
   ....: ['axis_coords', 'axis_types', 'data_unit', 'axis_units', 'title', 'axis_names', 'missing_value', 'data']

• Variables can be pulled from ferret into IPython with %ferret_getdata:

In [31]: import numpy as np
   ....: b = {}
   ....: b['name']='myvar'
   ....: b['name']='myvar'
   ....: x=np.linspace(-np.pi*4, np.pi*4, 500)
   ....: b['data']=np.sin(x)/x
   ....: b.keys()
Out[31]: ['data', 'name']
In [32]: %ferret_putdata --axis_pos (1,0,2,3,4,5) b
   ....: Message: b is now available in ferret as myvar

Explore notebook for examples.

Known issues and limitations

• Note that the pdf file produced by -p,--pdf option is not embedded in the notebook
• Limitations are the ones exposed from pyferret, the qualifier /pattern is not usable for example. Read http://ferret.pmel.noaa.gov/Ferret/documentation/pyferret/known-issues