numpy.float64 object is not iterable...but I'm NOT trying to

Question

I will provide the full code below, but the problem basically is this: I created a data structure like this: means = {ID1 : { HOUR1 : [AVERAGE_FLOW, NUMBER_OF_SAMPLES] ...}

I created AVERAGE_FLOW using np.mean(). I can do this:

print means['716353'][0][0] #OUT : 76.6818181818 

but when I run the second code when I want to :

means[row['ID']][i][0] 

I get: TypeError: 'numpy.float64' object is not iterable

Here are the codes, the first one is where I produce the means data, and the second where I am trying to create a list:

shunned=[]
means={}  #{ #DAY: [mean, number of samples]}
hour={}
for i in range(24):
    hour[i]=[]    
for station in stations:
    means[station]=copy.deepcopy(hour) 

for station in d:  
    for hour in range(24):
        temp=[]
        for day in range(1,31):
            if day in sb: #swtich between sa for all days and sb for business days
                try: #no entry = no counting in the mean, list index out of range, the    station has not hourly data to begin with
                    e = d[station][str(day)][hour][0]

                    if not e: # sometimes we have '' for flow which, should not be        counted
                        next
                    else:
                        temp.append(int(e))
                except IndexError:
                    if station not in shunned:
                        shunned.append([station,d[station]])
                    else:
                        next
        temp=np.array(temp)
        means[station][hour]=[np.mean(temp),len(temp)]

pprint.pprint(means)
print means['716353'][0][0] #OUT : 76.6818181818 






headers=['ID' , 'Lat', 'Lng', 'Link ID']+range(24)

csv_list=[]
meta_f.seek(0)
i=0
for row in meta_read:
    if i>100:
        break
    temp=[]
    if row['ID'] in stations:
        temp.append([row['ID'],row['Latitude'],row['Longitude'],' '])
        for i in range(24):
            temp.extend(means[row['ID']][i][0])
    csv_list.append(temp)
    i+=1

pprint.pprint(csv_list) #OUT:temp.extend(means[row['ID']][i][0]) TypeError: 'numpy.float64' object is not iterable

I tried str(np.means(temp)) in the first code thinking maybe it is because of numpy, but it actually gave me the first digit of my value! as if it is ITERATING through a string...could you please explain what is going on? thank you!

Answer 1

It looks like you're trying to extend a list with a scalar float variable. The argument to extend must be an iterable (i.e. not a float). From your first bit of code it looks like means[i][j][k] returns a float,

print means['716353'][0][0] #OUT : 76.6818181818

The problem is here,

temp.extend(means[row['ID']][i][0])

If you expect that means[i][j][k] will always be a single value and not a list you can use append instead of extend.

temp.append( means[row['ID']][i][0] )

An example to show the difference,

l = [i for i in range(10)]
l.extend( 99.0 )
TypeError: 'float' object is not iterable

this doesn't work b/c a float is not iterable

l.extend( [99.0] )
print l
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 99.0]

this works b/c a list is iterable (even a one element list)

l.append( 101.0 )
print l
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 99.0, 101.0]

append does work with a non-iterable (e.g. a float)