Perhaps an example will help,

mydict =
{’WILW1′: {’fx’: ‘8.1′, ‘obtime’: ‘2009-11-07 06:45:00′, ‘ob’: ‘6.9′},
‘GRRW1′: {’fx’: ‘12.8′, ‘obtime’: ‘2009-11-07 04:15:00′, ‘ob’: ‘6.7′},
‘NASW1′: {’fx’: ‘6.8′, ‘obtime’: ‘2009-11-07 06:30:00′, ‘ob’: ‘7.1′}
}

I would like to sort this by the values of the ‘ob’ key. In this case, this would become:

mysorteddic =
{’NASW1′: {’fx’: ‘6.8′, ‘obtime’: ‘2009-11-07 06:30:00′, ‘ob’: ‘7.1′},
‘WILW1′: {’fx’: ‘8.1′, ‘obtime’: ‘2009-11-07 06:45:00′, ‘ob’: ‘6.9′},
‘GRRW1′: {’fx’: ‘12.8′, ‘obtime’: ‘2009-11-07 04:15:00′, ‘ob’: ‘6.7′}
}

I’ve gone into the keys once and then again,

for k in keys:
for k2 in keys2:
if k2 == ‘ob’:
append to array or something…?
but it seems like I would have to keep track of the inner keys in an array, and then by that point I completely confuse myself. Thanks for your help!

–Jonathan

keys = car_details.keys()
keys.sort()
for k in keys: print car_details[k]

Thanks for your help!

car_details['Ford']['Mustang'] = 1
car_details['Toyota']['Corolla'] = 1

So in the above, I want to count the number for each one. In perl all I need to do is:

$car_details->{’Ford’}->{’Mustang’} ;
$car_details->{’Toyota’}->{’Corolla’} ;

Can you give me some idea?

Thanks,
Sam