Spread the love

Learn how to make beautiful GUI apps

in Python using the Kivy framework.

Comprehensive, for Kivy beginners, easy to follow.

Get the book here (PDF) or on Amazon:

ebook / paperback (black and white) / paperback (full color)

USING PANDAS MULTIINDEX FOR MULTIDIMENSIONAL DATA

We can use the MultiIndex for multidimensional data. Here’s our example from the previous part of the series:

In [2]:

import numpy as np
import pandas as pd

# the tuples used as indices
index = [('Company A', 2018), ('Company A', 2019),
        ('Company B', 2018), ('Company B', 2019),
        ('Company C', 2018), ('Company C', 2019)]

# the list of sales
sales = [125, 211,
        390, 455,
        475, 655]

# Let's create a Series object.
s = pd.Series(sales, index=index)

# Here's the multi-index.
m_index = pd.MultiIndex.from_tuples(index)

# Let's reindex the Series object.
s = s.reindex(m_index)
s

Out[2]:

Company A  2018    125
           2019    211
Company B  2018    390
           2019    455
Company C  2018    475
           2019    655
dtype: int64

Let’s convert the Series to a DataFrame. Let’s also make clear what sort of data the column contains:

In [3]:

s_df = pd.DataFrame({'sales': s})
s_df

Out[3]:

		sales
Company A	2018	125
Company A	2019	211
Company B	2018	390
Company B	2019	455
Company C	2018	475
Company C	2019	655

Now, what if we wanted to add one more dimension to the DataFrame? For example annual purchases? We can do it like so:

In [4]:

# Let's define the purchases as p:
p = [44, 72,
    81, 88,
    93, 87]

# Let's add the purchases column.
s_df2 = pd.DataFrame({'sales': s,
                     'purchases': p})

s_df2

Out[4]:

		sales	purchases
Company A	2018	125	44
Company A	2019	211	72
Company B	2018	390	81
Company B	2019	455	88
Company C	2018	475	93
Company C	2019	655	87

You can also use all the ufuncs with MultiIndex. Let’s calculate the difference between the sales and purchases for example:

In [5]:

b = s_df2['sales'] - s_df2['purchases']
b

Out[5]:

Company A  2018     81
           2019    139
Company B  2018    309
           2019    367
Company C  2018    382
           2019    568
dtype: int64

This will even look better if we unstack it:

In [6]: