getsizeof always 40 in python for each field in a csv -

this question has answer here:

• what difference between len() , sys.getsizeof() methods in python? [closed] 1 answer

so reading csv file row row. inside each row go field field , try size of each field in bytes suing .getsizeof. code below :

for row in reader:         temp1 = []         temp2 = []         if type(row[0]) inttype:             feed = feed + 1             print feed             # total number of columns in feed should 61.             # 61st column account last , after 60th column, blank.             #if len(row) == 61:             # total number of columns in feed should 61             field in row:                 if type(field) == 'int':                     field.encode('ascii', 'ignore')                     temp1.append(sys.getsizeof(field))                     temp2.append(str(field))                  else:                     field = [unicode(field)]                     #field = field.encode('ascii', 'ignore')                     temp1.append(sys.getsizeof(field))                     temp2.append(str(field)) 

for reason size foe every field accross rows coming out 40. idea why ?

first, this:

field.encode('ascii', 'ignore') 

… doesn't useful. doesn't change field; returns new bytes holding ascii-encoded version of field, don't store anywhere.

on top of that, call when field int, in case you'll attributeerror, because int objects don't have encode method.

so, clearly, you're hitting else case. , do?

well, makes 1-element list. you're not asking size of element, size of list. so, they'll same size.

as the docs explain:

only memory consumption directly attributed object accounted for, not memory consumption of objects refers to.

so, if want know size of list, plus 1 of elements? couple paragraphs down there's general-purpose solution:

see recursive sizeof recipe example of using getsizeof() recursively find size of containers , contents.

however, in case, there's simpler solution: sys.getsizeof(field) + sys.getsizeof(field[0]) do.

but note may not wanted anyway. example, if have 1000 rows, , 900 of them have value '0', you're going count u'0' 900 times… but really, you're storing 1 copy of u'0', 900 references it.