Cannot set an array element with a sequence

Question

I'm using the NumPy python library to run large-scale edits on a .csv file. I'm using this python code:

import numpy as np

def main():
    try:
        e,a,ad,c,s,z,ca,fn,ln,p,p2,g,ssn,cn,com,dob,doh,em = np.loadtxt('c:\wamp\www\_quac\carryover_data\SI\Employees.csv',delimiter=',',unpack=True,dtype='str')

        x=0
        dob = dob.split('/')
        for digit in dob:
            if len(digit) == 1:
                digit = str('0'+digit)
        dob = str(dob[2]+'-'+dob[0]+'-'+dob[1])

        doh = doh.split('/')
        for digit in doh:
            if len(digit) == 1:
                digit = str('0'+digit)
        doh = str(doh[2]+'-'+doh[0]+'-'+doh[1])

        for eID in e:
            saveLine=eID+','+a[x]+','+ad[x]+','+c[x]+','+s[x]+','+z[x]+','+ca[x]+','+fn[x]+','+ln[x]+','+p[x]+','+p2[x]+','+g[x]+','+ssn[x]+','+cn[x]+','+com[x]+','+dob[x]+','+doh[x]+','+em[x]+'\n'
            saveFile = open('fixedEmployees.csv','a')
            saveFile.write(saveLine)
            saveFile.close()
            x+=1


    except Exception, e:
        print str(e)

main()

dob and doh contain a string, e.g. 4/26/2012 and I'm trying to convert these to mysql friendly DATE forms, e.g. 2012-04-26. The error that is printed when I run this script is

cannot set an array element with a sequence

It does not specify a line and so I don't know what this really means. I'm pretty new to python; I've checked other questions with this same error but I can't make sense of their code. Any help is very appreciated.

also: perhaps this would be more easily done in excel? How could I do that? — khaverim
– khaverim, Commented May 13, 2014 at 5:52
Please remove the try: except clause and re-run your code to get the line and traceback for the error. — ebarr
– ebarr, Commented May 13, 2014 at 6:54
@ebarr doing so pointed me to the line with e,a,ad,c,s,z,ca,fn,ln,p,p2 etc. and cited the same exact error... — khaverim
– khaverim, Commented May 13, 2014 at 7:36
If you don't use a tuple unpack on the np.loadtxt command, does it execute without error? — ebarr
– ebarr, Commented May 13, 2014 at 10:28
@ebarr yeah, it still did. I managed to get things to work although it was somehwhat of a workaround. I'm posting the answer below. — khaverim
– khaverim, Commented May 13, 2014 at 18:21

alex · Accepted Answer · 2014-05-13 06:21:57Z

1

Try using zfill to reformat the date string so you can have a '0' before your '4'. (zfill pads a string on the left with zeros to fill the width.)

doh = '4/26/2012'
doh = doh.split('/')

for i, s in enumerate(doh):
    doh[i] = s.zfill(2)

doh = doh[2]+'-'+doh[0]+'-'+doh[1]

# result: '2012-04-26'

As for the cannot set an array element with a sequence it would be helpful to know where that is occurring. I'm guessing there is something wrong with structure of the array.

edited May 13, 2014 at 6:21

answered May 13, 2014 at 6:04

alex

6024 silver badges7 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

khaverim · Accepted Answer · 2014-05-15 20:50:40Z

Ok, to solve it I had to do a couple things. After removing the try-except commands, I found out that the error was on line 5, the line with e,a,ad,c,s etc. I couldn't eliminate the problem until I simply copied the 2 columns I wanted to focus on only and made a new program for dealing with those.

Then I had to create a .txt instead of a .csv because Excel auto-formats the dates and literally changes the values before I can even touch them. There is no way around that, I've learned. You can't turn the date-auto-format off. A serious problem with excel. So here's my solution for this NumPy script (it changes the first column and keeps the second the same):

import numpy as np
def main():
    dob,doh=np.loadtxt('temp.csv',
                             delimiter=',',
                             unpack=True,
                             dtype='str')


    x=0
    for eachDate in dob:
        if any(c.isalpha() for c in eachDate):
            newDate=eachDate
        elif (eachDate == ''):
            newDate=''
        else:
            sp = eachDate.split('/')
            y=0
            ndArray = ['','','']
            for eachDig in sp:
                if len(eachDig) == 1:
                    eachDig = str('0'+eachDig)
                if y == 0:
                    ndArray[0] = eachDig
                elif y == 1:
                    ndArray[1] = eachDig
                elif y == 2:
                    ndArray[2] = eachDig
                    newDate=str(ndArray[2]+'-'+ndArray[0]+'-'+ndArray[1])
                    y=0
                y+=1

        print eachDate+'--->'+newDate

        """creates a .txt file with the edited dates"""
        saveLine=str(newDate+','+doh[x]+'\n')
        saveFile=open('__newTemp.txt','a')
        saveFile.write(saveLine)
        saveFile.close()
        x+=1


main()

I then used Data->Import from text with "TEXT" format option in Excel to get the column into my .csv. I realize this is probably bulky and noobish but it got the job done :3

To summarize, I believe one of my columns in the original excel file was incompatible with numpy's loadtxt()

Collectives™ on Stack Overflow

Cannot set an array element with a sequence

2 Answers 2

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related