Histogram does not reflect quantum device results

Question

I created the following code to run a simple circuit through a quantum device twice. My purpose was to show the effects of noise with different arrangements of qubits when using ibmq_belem.

#Run circuit with two different qubit layouts
from qiskit import IBMQ, execute
#Load account
provider=IBMQ.load_account()
#Choose device
backend=provider.get_backend('ibmq_belem')
C=[]
for initial_layout in [[0,3],[2,4]]:
    job=execute(circuit, backend, initial_layout=initial_layout).result()
    c=result.get_counts()
    C.append(c)

followed by plot_histogram([C[0],C[1]],title="Comparing Noise") to create the histogram which give me ...

Which does not reflect any of the noise that I see in my results taken directly from IBMQ which shows the first and second runthroughs respectivly as ...

These are clearly not the identically perfect versions that my C set seems to have created. This makes my question; What have I done wrong in coding this that C is not accurately reflecting my counts?

Edit: I note that it has somehow decided to take the counts from when I ran this circuit on a simulator in a prior cell using the following code.

simulator=Aer.get_backend('aer_simulator')
result=execute(circuit,backend=simulator).result()
plot_histogram(result.get_counts(circuit)) 

This code had been run before I ever created the set C, so I did not expect it to interfere, but now I'm not sure how to work around this.

Answer 1

Figured it out! It was using my result label from the initial coding when I used the simulator and appending that twice to my set C. So in my code I have now changed c=result.get_counts() to c=job.get_counts() as that was the 'job' I wanted the results from! And now all is beautiful and full of noise!