###### tags: `python` `jupyter notebook` `replace` `re.sub`

# Python_string_manipulation

### Replace strings

---

#### Replace strings with 1 or multiple patterns. Note `re.sub`

import re
# Replace string ' (what ever)'
string='Bolivia (Plurinational State of)'
pattern1=' \\(.*\\)'
re.sub(pattern1,'',string) # working

# Remove any number that follow string
string2='Bolivia (Plurinational State of)20'
pattern2='[0-9]+'
re.sub(pattern2,'',string2) # working

# Replace string with 2 patterns
patterns= pattern1 + '|' + pattern2
re.sub(patterns,'',string2) # working

---

#### Replace strings in a column with 1 pattern

import pandas as pd
import numpy as np

# Import data
energy= pd.read_excel('D:/Now/workshops/20191013_coursera_Introduction-to-Data-Science-in-Python/course1_downloads/Energy Indicators.xls'
                     ,skiprows=17
                     ,names=column_names
                     ,nrows=227)
# Replace ... with NaN
energy = energy.replace('...', np.nan)             

---

#### Replace strings in a column with multiple patterns
* syntax: `DataFrame['column'].str.replace('pattern1|pattern2','replaced-string')`

# Replace strings
## regular expression is different in Python from bash (* in bash = .* in python)
## delete () and text within (e.g. )
## delete any number at any length that follows country names
energy['Country']= energy['Country'].str.replace(' \\(.*\\)|[0-9]+','')