Tableau : Convert ESRI shapes into Tableau Format

Click here for other Tableau related Blogs

Requirement is to convert data in ESRI Map shapes into Tableau format.

Tools Required

1. Quantum GIS - http://www.qgis.org/en/site/
2. Python 
3. Download the below Python Code - https://github.com/venkat-vs-id/python

import csv
import re
import sys

#def convertESRI2TableauFormat( p_inputfile, p_outputfile):

l_outfileName = 'c:/ESRI_suburb_tab_format.csv'  # make changes here..
l_ESRIFileName ='c:/ESRI_suburb.csv'             # make changes here..

rdx=0
csv.field_size_limit(sys.maxsize)
outfile = open( l_outfileName, 'w', newline='')
csvWriter = csv.writer( outfile, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)

with open( l_ESRIFileName, 'r') as csvfile:

 csvR = csv.reader( csvfile, delimiter=',', quotechar='"')
 for row in csvR:
    rdx += 1           
    ci=0
    rowValues = []
    rowValues.append(row[1]);# ssc_code
    rowValues.append(row[2]);# ssc_name
    rowValues.append(row[4]);# state_code
    rowValues.append(row[5]);# confidence
    rowValues.append(row[6].strip());# AREA_SQKM
    rowValues.append('');# place holder for lat
    rowValues.append('');# place holder for long
    rowValues.append('');# place holder for polygon
    rowValues.append('');# place holder for point
      
    if ( rdx==1): # first row add header col names

       rowValues[5]=('Latitude');# 
       rowValues[6]=('Longitude');#
       rowValues[7]=('Polygon_ID');#
       rowValues[8]=('Point_ID');#
       csvWriter.writerow( rowValues);
    else:        
       for col in row:

           ci += 1
           LatLonStrArr=[]

           if (ci==1): # first col has the polygon values.

               polygonArr = []
               polygonArr = re.findall('(\((?:-?\d+\.\d+\s-?\d+\.\d+,?)+\))', col)
               polyx =0
                     
               for px in polygonArr:

                   polyx += 1
                   LatLonStrArr = []
                   LatLonStrArr = re.findall('(-?\d+\.\d+\s-?\d+\.\d+)', px)

                   if( len( LatLonStrArr) != 0 ):

                       pointx  = 0
                       pointxx = 0
                       for lx in LatLonStrArr:

                           pointx += 1
                           resArr = re.findall( '(-?\d+\.\d+)', lx)
                           #print('>>>> lx='+lx + "/"+ str(len(resArr)))
                           rowValues[5] = resArr[1] # lat
                           rowValues[6] = resArr[0] #long
                           rowValues[7] = polyx

                           if( (pointx ==1) | (pointx == len(LatLonStrArr))| (len(LatLonStrArr)<=10)| (pointx % 10 == 0) ):
                              pointxx += 1
                              rowValues[8] = pointxx
                              csvWriter.writerow( rowValues);
   
    #print('ROW========================='+ str(rdx))
    #for rx in rowValues:
    #   print('val=' + str(rx))
    
    #if (rdx >=100):
       #break
            
csvfile.close()
outfile.close()

For the sake of this exercise I downloaded the census data from ABS website ( free registration required).

Step to convert ESRI data into Tableau Format

1. Download your ESRI data. and it will look like this 
2. Open Quantum GIS
3. From the top menu -> new project
4. From the top menu -> Layer -> "Add Vector Layer" ( in this select the .shp file downloaded in step 1) 
   
5. Now you will be able to see the From maps and Shapes from the <filename>.shp file
6. Right-click and select "save as"
   
7. Make sure you save as with the Layer Option as GEOMETRY=AS_WKT 
   
8. Now you have exported the ESRI  shp file into a CSV
9. Use the Python code from above, and make the changes to the 2 variables 
1. l_outfileName
2. l_ESRIFileName
10. Run the above py code- you will get the CSV file readable by Tableau
11. If you dont know how to use that CSV in Tableau - Look into this Blog - Tableau Maps - Custom shapes / polygon - Basics