python中两个值之间的差异问题

Question

i'm starting learning python and during an analysis i need to computer every value in a column with any other value in the same column with 100005 entry. I was thinking about using a code like this:

for ts in u.trajectory:
  distanza5=distance_array(coppia5.positions, coppia5.positions)
  k=pd.DataFrame(distanza5)
  valore=k.iat[0,1]
  list.append(valore) 

step1 = pd.concat([dist1,dist2], axis=1)

step2 = pd.concat([step1,dist3], axis=1)

step3 = pd.concat([step2,dist4], axis=1)

step4 = pd.concat([step3,dist5], axis=1)

 for i in step4.index:
   first=step4.iat[i,0]  
   y=abs(first)
   for n in range (0,100004):
     x=abs(step4.iat[n,0])
     value[n]=y-x
     list.append(value[n])
   diff[i]=pd.DataFrame(list)
   del list

But when i try to run it i get this error: TypeError: 'numpy.float64' object does not support item assignment

The problem is given from the value[n] line.What can i do?

Thanks for your help

Answer 1

For the second part of your code, the problem seems to be caused by the initialization of your value variable. This variable probably is initialized as a single float, for example by writing value = k.iat[0,1] . In order as intended you have to use a list or some indexable data structure. For example you can get rid of the value variable as shown in the following code:

 for i in step4.index:
   first=step4.iat[i,0]  
   y=abs(first)
   list = []
   for n in range (0,100004):
     x=abs(step4.iat[n,0])
     list.append(y-x)
   diff[i]=pd.DataFrame(list)

To solve the declaration problem of the diff variable, you can add this line diff = pd.DataFrame() before the for i in step4.index line and also change the last line of the code in this way:

diff = pd.DataFrame()
for i in step4.index:
   first=step4.iat[i,0]  
   y=abs(first)
   list = []
   for n in range (0,100004):
     x=abs(step4.iat[n,0])
     list.append(y-x)
   diff.loc[:,i] = pd.Series(list)

This new last line adds the entire list in a new column inside the diff dataframe created