Question

I have 2 nested dictionaries in Python that have this format:

1166869: {'probL2': '0.000', 'probL1': '0.000', 'pronNDiff_site': '1.000', 'StateBin': '0', 'chr': 'chrX', 'rangehist': '59254000-59255000', 'start_bin': '59254000', 'countL2': '4', 'countL1': '0'} 

1166870: {'probL2': '0.148', 'probL1': '0.000', 'pronNDiff_site': '0.851', 'StateBin': '0', 'chr': 'chr2', 'rangehist': '59254000-59255000', 'start_bin': '59255000', 'countL2': '5', 'countL1': '15'} 

1166871: {'probL2': '0.000', 'probL1': '0.000', 'pronNDiff_site': '1.000', 'StateBin': '0', 'chr': 'chrY', 'rangehist': '59290000-59291000', 'start_bin': '59290000', 'countL2': '1', 'countL1': '2'}

where 1166869, 1166870 and 1166871 represent a line in a file from where I read the data, and the rest of the keys are the data itself.

Now I want to make a list where I store all the different values in the key “chr” because there are some repeated ones.

How can I go through the dictionary and make the comparison between the 2 values? This code is not working:

for k in range(len(file_dict)):
    for j in range(len(file_dict)-1):
        if (file_dict[j]["chr"] != file_dict[k]["chr"]):
            list_chr.append(file_dict[j]["chr"])

Answer 1

Use a set, and just all the items in one go:

chr = { v['chr'] for v in file_dict.itervalues() }

This uses a set comprehension to generate your set in one line of code.

Set comprehensions were introduced in Python 2.7; in earlier versions use:

chr = set(v['chr'] for v in file_dict.itervalues())

In Python 3, you’d need to replace .itervalues() by .values().

Your own code doesn’t work because python dictionaries are not lists; you don’t retrieve values by index, but by key. You’d have to change it to:

for key in file_dict:
    for other_key in file_dict:
        if key == other_key:
            continue

        if file_dict[key]['chr'] != file_dict[otherkey]['chr']:
            list_chr.append(filed_dict[key]['chr'])

but that is really inefficient, not to mention incorrect.