python probability density function

Question

I can’t create a continuous probability density function for a random variable to plot it over the histogram.

val = df[['posição', 'num']].groupby('num').agg(lambda x: list(set(x))).reset_index()
cont = 0
fprob=[]
for i in val.index:
    cont += len(val.query('index=={}'.format(i))
    fprob.append(len(val.query('index=={}'.format(i))/(cont*(len(df)/len(var)))

That’s what I tried, where the val keeps the positions where the value one repeats within df. Then I tried to search for each value how many times it repeats and divide it by the amount of values that have already been tested times the proportion that each value represents of df. I don’t know any function that already provides me the probability density function, so I tried it like this.