用于创建2D数组形式2 2D数组的python for循环

Question

I want to use python for getting vibrational partition function at different temperature. I have two 1D array like: freq=[100,150,200] (This is actually read form a .dat file) and Temp=np.arange(273.,973.,25) . Now, qvib=product over freq for a particular temperature (1/1-exp(-h*nu/kb/T)) . My for loop is:

for i in range(len(T)):
 for j in range(len(freq)):
  q[i][j]=1/(1-np.exp(-h*freq[j]/kb/T[i]))
   Q[i]=np.prod(q[i][:])

I am using python for two to three days only. I even don't know all the in line operation of python. I have two question: 1) Can i do it without using for loop (like using only sum or no.prod) 2) My for loop is not working. I searched on internet but couldn't found any good for loop reference for this type of mathematical operation. Thanks in advance for your kind help.

Answer 1

You need to at least define Q as an empty array the same length as your temperature array before you can assign to it.:

Q = np.empty(len(T))

Better is:

for i in range(len(T)):
    q = 1/(1-np.exp(-h*freq/kb/T[i]))     
Q = np.prod(q)

But the really nice thing about NumPy is the ability to do away with Python loops altogether:

Q = np.prod(1/(1-np.exp(-h*freq[:, None]/kb/T)), axis=0)