emperical rule in Python

from statsmodels.distributions.empirical_distribution import ECDF
import seaborn as sns
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import random
import warnings
warnings.filterwarnings('ignore')

random.seed(1738)

mu = 7
sigma = 1.7
Observations = [random.normalvariate(mu, sigma) for _ in range(100000)]

sns.distplot(Observations)
plt.axvline(np.mean(Observations) + np.std(Observations), color="g")
plt.axvline(np.mean(Observations) - np.std(Observations), color="g")

plt.axvline(np.mean(Observations) + (np.std(Observations) * 2), color="y")
plt.axvline(np.mean(Observations) - (np.std(Observations) * 2), color="y")


pd.Series(Observations).describe()  # mean=7.001202; std=1.701952;


SampleA = random.sample(Observations, 100)
SampleB = random.sample(Observations, 100)
SampleC = random.sample(Observations, 100)

fig, ax = plt.subplots()
sns.distplot(SampleA, ax=ax)
sns.distplot(SampleB, ax=ax)
sns.distplot(SampleC, ax=ax)


ecdf = ECDF(Observations)
plt.plot(ecdf.x, ecdf.y)
plt.axhline(y=0.025, color='y', linestyle='-')
plt.axvline(x=np.mean(Observations) -
            (2 * np.std(Observations)), color='y', linestyle='-')
plt.axhline(y=0.975, color='y', linestyle='-')
plt.axvline(x=np.mean(Observations) +
            (2 * np.std(Observations)), color='y', linestyle='-')