作为解决问题的一般方法，你应该试着把你的问题分解成更小的可解决的块，以便更好地理解它。
解决这个问题的一个很好的起点如下

import pandas as pd
import numpy as np

random_nums = np.array([1, 9, 1, 3, 0, 3, 6, 0, 2, 0])

df = pd.DataFrame({"numbers": range(min(random_nums), max(random_nums) + 1)})

df["total"] = df["numbers"].map(lambda x: np.count_nonzero(random_nums == x))

df["first_occurrence"] = df["numbers"].map(lambda x: np.argmax(random_nums == x))

df["delta"] = np.nan

# Delta only makes sense if the number of duplicates are atleast 2
df.loc[df.total >= 2, "delta"] = df.loc[df.total >= 2, "numbers"].map(
    lambda x: np.ptp(np.where(random_nums == x))  # Peak to peak difference of indices
)

print(df)

输出

numbers  total  first_occurrence  delta
0        0      3                 4    5.0
1        1      2                 0    2.0
2        2      1                 8    NaN
3        3      2                 3    2.0
4        4      0                 0    NaN
5        5      0                 0    NaN
6        6      1                 6    NaN
7        7      0                 0    NaN
8        8      0                 0    NaN
9        9      1                 1    NaN