您可以用途：

# convert to datetime
s = pd.to_datetime(df['Time'])

# identify the center
center = s[df['event'].eq('Yes')].iloc[0]

# identify if the date is before/center/after
group = (np.sign(s.sub(center).dt.days.astype(int))
           .map({-1: 'pre_center', 0: 'center', 1: 'post_center'})
        )

# aggregate
out = df.groupby(group)['value'].agg(average_value='sum')

输出量：

Range  average_value
0       center             29
1  post_center             61
2   pre_center             42

如果要包括阈值：

s = pd.to_datetime(df['Time'])

center = s[df['event'].eq('Yes')].iloc[0]

diff = pd.DateOffset(months=2)
m1 = s.between(center-diff, center)
m2 = s.between(center, center+diff)
group = np.select([m1&m2, m1, m2], ['center', 'pre_center', 'post_center'], np.nan)

out = (df.groupby(group)['value']
         .agg(average_value='sum')
         .drop('nan', errors='ignore')
         .rename_axis('Range').reset_index()
      )

输出量：

Range  average_value
0       center             29
1  post_center             30
2   pre_center             22