Pandas transpose rows to columns

问题

I am writing the result of an sql query into an excel sheet and attempting to transpose rows into columns but cannot seem to get Pandas to budge, there seems to be an conundrum of some sort with excel. I have done my googlefu and have looked at:

how to switch columns rows in a pandas dataframe

How to convert rows values in dataframe to columns labels in Python after groupby?

Convert row to column in Python Pandas

Python Pandas: Convert Rows as Column headers

None seems to work. Please advise on this.

import psycopg2
import pandas as pd
import xlsxwriter

try:
    conn = psycopg2.connect(private stuff cannot be shared)
except:
    print ("I am unable to connect to the database")

cursor = conn.cursor()

writer = pd.ExcelWriter("Z:/AWS/SQLQueries/Phoebe's Request.xlsx",engine = 'xlsxwriter')

query20 = """SELECT 2 AS rowtype
 , source AS "TrafficTypes_Name"
 , COUNT(source) AS "Traffic"
 , to_char(week,'MM/dd/yyyy') AS "Week_Ending"
FROM amazon.tracker
where project_id = 'PCR'
GROUP 
BY source
 , to_char(week,'MM/dd/yyyy')
UNION ALL
SELECT 1 
 , 'Visitor Center Walk-ins'
 , COUNT(source)
 , to_char(week,'MM/dd/yyyy') as week 
FROM amazon.tracker
where project_id = 'PCR'
GROUP 
BY to_char(week,'MM/dd/yyyy')
ORDER 
BY "Week_Ending"
 , rowtype"""

cursor.execute(query20)

result = cursor.fetchall()

first = pd.DataFrame(result, columns = ["rowtype","TrafficTypes_Name","Traffic","Week_Ending"])

first.drop(first.columns[0],axis=1, inplace = True)

first.pivot(index = 'Week_Ending', columns = 'TrafficTypes_Name' , values = 'Traffic' )

first.to_excel(writer, sheet_name = 'Visitor Traffic',index = False)

print ("Query 20 Created")

writer.save()

Excel sheet:

TrafficTypes_Name       Traffic          Week_Ending

Visitor Center Walk-ins   18             01/01/2017
Resident Referral          1             01/01/2017
Community Website          1             01/01/2017
Realtor                    1             01/01/2017
Other Website              1             01/01/2017
Social Media               1             01/01/2017
Builder                    3             01/01/2017
Drive-by                   10            01/01/2017

Requesting:

Week_Ending    Visitor Center Walk-ins       Resident Referral         Community Website            Realtor  ....................
01/01/2017          18                        1                           1                             1  .........................

回答1:

Pivot would work here

df1 = df.pivot(index='Week_Ending', columns='TrafficTypes_Name', values='Traffic')


TrafficTypes_Name   Builder CommunityWebsite    Drive-by    OtherWebsite    Realtor ResidentReferral    SocialMedia VisitorCenterWalk-ins
Week_Ending                             
01/01/2017          3       1                   10          1               1       1                   1            18

来源：https://stackoverflow.com/questions/43639978/pandas-transpose-rows-to-columns

标签

python

pandas

psycopg2