Pytrends to get trending search suggestions for Keywords

Pytrends is the unoffical API for Google Trends.
If Pytrends is not installed, this error message will be displayed.
ModuleNotFoundError: No module named 'pytrends'
How to install Pytrend? In your command prompt enter this line.

pip install pytrends

Installing Pytrend in Colab

Installing Pytrends in google colab platform

!pip install pytrends

Learn more about Google Colab platform.
Getting today's trending search using trending_searches()

from pytrends.request import TrendReq
pytrends = TrendReq()
df = pytrends.trending_searches(pn='india')
print(df)

Output

0                   AHMEDABAD
1                       Putin
2          Shani Jayanti 2022
3                  Today news
4           Nottingham Forest
5                          IP
6                      Legend

Check the output at Google Trends

realtime_trending_searches()

We can use pn='US' , pn='IN' etc. for country specific search. List of countries with code is available here.

df = pytrends.realtime_trending_searches(pn='CA')

suggestions

By using key word we can get suggestions from google trend.

my_list=pytrends.suggestions('Digital Marketing') 
for val in my_list:
    print(val['type'],'->',val['title'])

Related queries

kw=['Digital Marketing','Social media marketing','content marketing']
pytrends.build_payload(kw_list=kw,timeframe='today 1-m',geo='IN')
pytrends.related_queries()

Managing time frame:

today 1-m : From today previous one month. Here month can only take values 1,3 or 12 only.
now 1-d : Last one day only, it can take value 1 or 7 only.
now 1-H : Last one hour, it works for 1 and 4 hours only.
today 5-y : From today last 5 years
'2021-12-19 2022-01-10' YYYY-MM-DD YYYY-MM-DD for specific period.
'2022-02-06T10 2022-02-12T07' With datetime using UTC

Interest by region

kw=['Digital Marketing','Social media marketing','content marketing']
pytrends.build_payload(kw_list=kw,timeframe='today 5-y',geo='IN')
#pytrends.build_payload(kw_list=kw,timeframe='2022-02-06T10 2022-02-12T07',geo='US')
pytrends.interest_by_region()

Historical Hourly Interest

kw=['Python','PHP','JavaScript','HTML','MySQL']
df=pytrends.get_historical_interest(kw, year_start=2020, 
	month_start=1, day_start=1, hour_start=0, 
	year_end=2020, month_end=2, day_end=1, hour_end=0, 
	cat=0, geo='', gprop='', sleep=0)
#df.plot(subplots=True, figsize=(20, 12))

Output

                 date  Python  PHP  JavaScript  HTML  MySQL  isPartial
0 2020-01-01 00:00:00      32   17           8    13      4      False
1 2020-01-01 01:00:00      31   15           8    14      5      False
2 2020-01-01 02:00:00      30   18           8     8      6      False
3 2020-01-01 03:00:00      33   22           7    12      4      False
4 2020-01-01 04:00:00      33   21          10    15      6      False

We need to resample above data to get data over a period. Above data is an interval of one hour so let us take the daily average value or monthly average value. Before using resample() we need to convert the date column to Datetime.

import pandas as pd
kw=['Python','PHP','JavaScript','HTML','MySQL']
df=pytrends.get_historical_interest(kw, year_start=2020, 
	month_start=1, day_start=1,  
    year_end=2021, month_end=1, day_end=10, 
    cat=0, geo='', gprop='', sleep=0).reset_index()
#print(df.columns)
df['date']=pd.to_datetime(df['date'])
df=df.resample('M', on='date').mean()
print(df)

Output

               Python        PHP  JavaScript       HTML      MySQL  isPartial
date                                                                         
2020-01-31  57.078877  27.098930   21.644385  27.423797  11.402406        0.0
2020-02-29  63.112857  27.844286   23.637143  30.468571  12.457143        0.0
2020-03-31  65.108289  28.981283   24.279412  31.517380  12.831551        0.0
2020-04-30  74.303448  29.267586   26.544828  33.942069  13.335172        0.0
2020-05-31  70.735294  28.426471   25.195187  32.414439  12.677807        0.0
2020-06-30  68.388122  29.237569   25.604972  31.824586  12.969613        0.0
2020-07-31  67.160214  28.958611   25.068091  30.510013  12.595461        0.0
2020-08-31  64.627005  27.794118   23.929144  29.491979  11.782086        0.0
2020-09-30  64.201379  25.779310   22.740690  29.801379  11.307586        0.0
2020-10-31  64.545455  25.631016   21.600267  29.926471  10.553476        0.0
2020-11-30  67.218232  27.556630   21.798343  30.788674  11.160221        0.0
2020-12-31  64.389853  27.753004   21.587450  29.837116  10.712951        0.0
2021-01-31  60.399083  27.871560   19.917431  27.233945   9.545872        0.0

Let us create a plot

import pandas as pd
kw=['Python','PHP','JavaScript','HTML','MySQL']
df=pytrends.get_historical_interest(kw, year_start=2020, 
          month_start=1, day_start=1,  
          year_end=2021, month_end=1, day_end=10, 
          cat=0, geo='', gprop='', sleep=0).reset_index()
df['date']=pd.to_datetime(df['date'])  # create date column 
df=df.resample('M', on='date').mean() # resample using monthly average
df.reset_index(inplace=True) # remove index from date column
df.drop(labels='isPartial',axis=1,inplace=True) # delete coloumn
print(df.columns)
df.plot(figsize=(10,5),x='date')

Note that numbers ( in Y Axis ) are scaled on a range of 0 to 100 based on a topic’s proportion to all searches on all topics.

Creating CSV or Excel file

We can collect data and store in a csv ( comma separated values ) or store in Excel file.

import pandas as pd
from pytrends.request import TrendReq
pytrends = TrendReq()
kw=['Python','PHP','JavaScript','HTML','MySQL']
df=pytrends.get_historical_interest(kw, year_start=2020, 
          month_start=1, day_start=1,  
          year_end=2021, month_end=1, day_end=10, 
          cat=0, geo='', gprop='', sleep=0).reset_index()
#df.to_excel('D:\my_data\my_trends.xlsx',index=False) # excel file
df.to_csv('D:\my_data\my_trends.csv',index=False) # csv  file

Storing in Database

We can use to_sql() to store data in MySQL Database.

from sqlalchemy import create_engine
my_conn = create_engine("mysql+mysqldb://userid:pw@localhost/my_db")
df.to_sql(con=my_conn,name='trends',if_exists='append')

Pandas list Python- Tutorials

Subscribe to our YouTube Channel here

Pytrends

Installing Pytrend in Colab

realtime_trending_searches()

Top searches for the past years

suggestions

Related queries

Managing time frame:

Related Topics

Interest by region

Historical Hourly Interest

Creating CSV or Excel file

Storing in Database