Nik piepenbreier datagy. Io 31 tips and tricks for super stars

Let's get started! 
We’ll kick things off by importing our required libraries: 
In [1]: 
import
​
pandas 
​
as
​
pd 
import
​
numpy 
​
as
​
np
Nik Piepenbreier - 
​
datagy.io
 

---

6 
Tip #1: Scrape a Website with One 
Function 
Easily scrape a web table (or multiple web tables) using Pandas. 
Use the Pandas read_html() function to easily extract data from web tables. 
In [2]: 
website = 
"https://en.wikipedia.org/wiki/All-time_Olympic_Games_medal_table"
df_list = pd.read_html(website) 
df = df_list[
​
2
​
] 
df
Out[2]: 
Team (IOC code)
№ Summer
№ Winter
№ Games
0
Albania (ALB)
8
4
12
1
American Samoa (ASA)
8
1
9
2
Andorra (AND)
11
12
23
3
Angola (ANG)
9
0
9
4
Antigua and Barbuda (ANT)
10
0
10
...
...
...
...
...
74
Republic of China (ROC) [ROC]
3
0
3
75
Saar (SAA) [SAA]
1
0
1
76
North Yemen (YAR) [YAR]
2
0
2
77
South Yemen (YMD) [YMD]
1
0
1
78
Refugee Olympic Team (ROT) [ROT]
1
0
1
79 rows × 4 columns 
The function returns a list of dataframes on the webpage. 
Slice the list to find your table and assign it to a dataframe. 
Nik Piepenbreier - 
​
datagy.io
 

---

7 
Tip #2: Reverse a List in Python 
To reverse the values of a list, use the negative indexer twice in a row. 
In [3]: 
original_list = [
​
1
​
,
​
2
​
,
​
3
​
,
​
4
​
,
​
5
​
] 
reversed_list = original_list[::
​
-1
​
] 
reversed_list
Out[3]: 
[
​
5
​
, 
​
4
​
, 
​
3
​
, 
​
2
​
, 
​
1
​
]
This uses ​extended slices​ to reverse the list. 
Extended slices add a third "step" argument to a slice. 
By using a step of -1, Python returns the list in reverse order. 
This happens because it starts at the -1 (the last value), then steps another -1 
to -2, until position 0 is hit. 
Nik Piepenbreier - 
​
datagy.io
 

---

8 
Tip #3: Speed Up Loading Dataframes 
by Assigning Data Types 
Pandas has 7 different data types: 
1. Object (object - including strings and mixed) 
2. Integer (int64) 
3. Float (float64) 
4. Boolean (bool) 
5. Datetime (datetime64) 
6. Time delta (timedela[ns]) 
7.
Categorical (category) (Check out ​Tip #9​ to learn more about 
categorical data types) 
Each column in a dataframe can only have one data type. 
Pandas will attempt to identify the data type for each column when you 
import it by scanning across the values in that column. 
To speed up the import, if you know the column data types of specific 
columns, you can include them as an argument in the import function. 
In [4]: 
df = 
pd.read_csv(
​
'https://raw.githubusercontent.com/datagy/pivot_table_
pandas/master/select_columns.csv'
​
dtype={
​
'Name'
​
:str, 
​
'Age'
​
:int, 
'Height'
​
:str, 
​
'Score'
​
:int, 
​
'Random_A'
​
:int, 
​
'Random_B'
​
:int, 
'Random_C'
​
:int, 
​
'Random_D'
​
:int}) 
df
Out[4]: