#!/usr/bin/env python
# coding: utf-8

# # Exporting data
# 
# There are fundamentally two different data types:
# 
# - spatially averaged data (as in the `odt` [OOMMF Data Table] file)
# - spatially resolved data (as in the `omf`, `omv` files)

# ## Access the raw OOMMF output files
# 
# When Jupyter OOMMF needs a micromagnetic calculation to be carried out, it is writing a `mif` file, and then asks OOMMF to execute that miffile, and export data to the file system.
# 
# It it thus possible to read the output files from the disk as one would normally with OOMMF. This allows re-use of all existing tools to analyse OOMMF data (including muview, ovf2vtk, ...)
# 
# Example

# In[1]:


import oommfc as oc
system = oc.examples.macrospin()
td = oc.TimeDriver()
td.drive(system, t=0.1e-9, n=5)


# From the last output line, we can see that the OOMMF data is stored in the directory `example-macrospin`. We can display the content:

# In[2]:


get_ipython().system('ls example-macrospin/')


# ## Exporting the OOMMF Data Table
# The file `example-macrospin.odt` is automatically read and available as a pandas data table in the `system.dt` object. We can thus use all the export functionality that pandas supports. For example

# In[3]:


# display beginning of data table in notebook
system.dt.head()


# ### MS Excel, html, latex, json

# In[4]:


system.dt.to_excel("data-odt.xlsx")


# In[5]:


system.dt.to_html("data-odt.html")


# In[6]:


system.dt.to_latex("data-odt.tex")


# In[7]:


system.dt.to_json("data-odt.json")


# ### Numpy arrays

# Given a pandas table, we can extract each column as a (Pandas) Series:

# In[10]:


t_series = system.dt['t']
t_series


# Or convert this into a numpy array:

# In[11]:


import numpy as np
t_array = np.array(system.dt['t'])
mx_array = np.array(system.dt['mx'])


# for example to plot the data in some custom way:

# In[12]:


get_ipython().run_line_magic('matplotlib', 'inline')
import matplotlib.pyplot as plt
fig, axes = plt.subplots()
axes.plot(t_array, mx_array)


# To see which data columns are available in the table, we can use:

# In[13]:


system.dt.columns


# or

# In[14]:


list(system.dt.columns)