Introduction to pandas & NumPy

pandas is a popular data analysis library in Python. You can think of pandas as a back-end Excel tool that can be customized in order to deal with raw data more easily. pandas deals with 1D and 2D (dimensional) arrays.

1. NumPy

import numpy as np

np.array([1,3,5])

array([1, 3, 5])

np.array([[1,3,5], [2,4,6]])

array([[1, 3, 5],
       [2, 4, 6]])

np.zeros([3,6])

array([[0., 0., 0., 0., 0., 0.],
       [0., 0., 0., 0., 0., 0.],
       [0., 0., 0., 0., 0., 0.]])

a = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])
a.shape

(3, 4)

a = np.array(['a', 'b', 3, 4])

print(a[1:3])

['b' '3']

b = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])
print(b)

[[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]]

print(b[0:2, 1:3])

[[2 3]
 [6 7]]

print(b[:, 1:3])

[[ 2  3]
 [ 6  7]
 [10 11]]

print(b[1:, :])

[[ 5  6  7  8]
 [ 9 10 11 12]]

print(b[:, :])

[[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]]

Let’s use NumPy for calculation.

n1 = np.array([3, 6, 9])

print(n1)
print(n1 + 5)
print(n1 - 5)
print(n1 * 5)
print(n1 / 5)

[3 6 9]
[ 8 11 14]
[-2  1  4]
[15 30 45]
[0.6 1.2 1.8]
[3 1 4]

n2 = np.array([[1,2,3,4,5], [6,7,8,9,10]])
print(np.max(n2))
print(np.min(n2))
print(np.mean(n2))
print(np.median(n2))
print(np.var(n2))
print(np.std(n2))

10
1
5.5
5.5
8.25
2.8722813232690143

2. pandas

pandas has two data structures: Series for 1D data and DataFrame for 2D data.

import pandas as pd

# Series

s1 = pd.Series({'Boffin': 100, 'Bella': 90, 'John': 85})
s1.name = 'Literature'
print(s1)

Boffin    100
Bella      90
John       85
Name: Literature, dtype: int64

# DataFrame

pd.DataFrame({
    'age':[10, 42, 21]
})

	age
0	10
1	42
2	21

{
    'age':[10, 42, 21]
}

{'age': [10, 42, 21]}

{
  'age':[10, 42, 10],
  'name': ['Boffin', 'Bella', 'John'],
  'gender': ['male', 'female', 'male']
}

{'age': [10, 42, 10],
 'gender': ['male', 'female', 'male'],
 'name': ['Boffin', 'Bella', 'John']}

pd.DataFrame({
  'age':[10, 42, 21]
}).join(
pd.DataFrame({
  'name': ['Boffin', 'Bella', 'John']
})
)

	age	name
0	10	Boffin
1	42	Bella
2	21	John

pd.DataFrame({
  'age':[10, 42, 21]
  }).join(
      pd.DataFrame({
        'name': ['Boffin', 'Bella', 'John']
        })
    ).join(
      pd.DataFrame({
        'city': ['Bryan', 'Seoul', 'London']
      })
)

	age	name	city
0	10	Boffin	Bryan
1	42	Bella	Seoul
2	21	John	London

pd.DataFrame({
  'age':[10, 42, 21]
}).append(
pd.DataFrame({
  'name': ['Boffin', 'Bella', 'John']
})
)

	age	name
0	10.0	NaN
1	42.0	NaN
2	21.0	NaN
0	NaN	Boffin
1	NaN	Bella
2	NaN	John

pd.DataFrame({
  'age':[10, 42, 21],
  'name': ['Boffin', 'Bella', 'John']
}).merge(
pd.DataFrame({
  'name': ['Boffin', 'Bella', 'John'],
  'city': ['Bryan', 'Seoul', 'London']
}), on=['name']
)

	age	name	city
0	10	Boffin	Bryan
1	42	Bella	Seoul
2	21	John	London