blog

import numpy as np
import pandas as pd
    
data = pd.read_csv("filename")
data.head(6)
    
#select one column
data['Column name 1']
# select multiple columns
df = data[['Column1', 'Column2']]
# select rows
data[3:10]
    
# conditional selection
data[data['Column1']<500]

# check type of data
type(data)
# convert to dataframe
df = pd.DataFrame(data)
# remove first row
df.drop(df.index[0], inplace=True)
# remove first two rows
df.drop(df.index[:2], inplace=True)
# drop column by name
df.drop(['Column1'], axis=1, inplace=True)

# select all entries from index 2 onwards
data[2:]
# select entries at even index locations
data[::2]
# specify the ranges of rows and columns
data.iloc[0:8, 0:7]
# select all columns for rows of index values 0 and 10
data.loc[[0,10], :]
# isolate rows and columns
data[['Column1', 'Column2']][4:7]
# isolate by row names
data[data.Countries == "Italy"]
# isolate by two row labels
data.loc[data['Countries'].isin(['Italy', 'Germany'])]
# select rows with multiple conditions
data[(data.Countries == 'India') & (data.Degree == 'Vulnerable')]
# change column name
df = data[:] # copy data
df.rename(columns={'old name' : 'new name'}, inplace = True)

# sort in descending order by an attribute
large = data.sort_values(by='Number of speakers', ascending=False)
# group data
data.groupby(['Column1', 'Column2'])  # not the same as gorup by in MR, but while doing stats it shows the stats for grouped data
# get the grouped stats
data.groupby(['Column1']).count()
# group one column by anther column
# the result is row_number1 column1_1
#               row_number2 column1_2 etc
x = data['Column1'].groupby(data['Column2'])
# lambda functions
# identify which of the languages have fewer than 5k speakers
# create a new column, if Number of speakers is less than 5000, it will be true, otherwise false
data['view'] = data['Number of speakers'].apply(lambda x: x<= 5000)

# tabulate data on basis of two columns
# something like the horizontal are attributes of column1, vertical are attributes of columm2, each cell is the count
# of number of joint count
x = pd.crosstab(data['Column1'], data['Column2'], margins=True)