用历史list 判断 一个value
from scipy import stats
history = [0.05, 0, 0, 0.07, 0.01]
new_score = 0.01
result = stats.ttest_1samp(history, new_score)
print(result)print结果：TtestResult(statistic=1.114772822866588, pvalue=0.32740571539283314, df=4)
用历史list 判断 另一个list
from scipy import stats
history = [0.05, 0, 0, 0.07, 0.01]
new_score_list = [0.01, 0.01]
result = stats.ttest_ind(history, new_score_list)
print(result)print结果：Ttest_indResult(statistic=0.6662041853993044, pvalue=0.5347812362215895)
pvalue 越小，说明实验 和历史差的多，置信，超过波动
pvalue 越大，说明实验 和历史差不多，不置信，为波动