对于组中所有值的绝对匹配，可以使用groupby.transform('all')：

out = df[df['STRATEGY'].eq('SELL').groupby(df['BOND']).transform('all')]

字符串
groupby.all：

s = df['STRATEGY'].eq('SELL').groupby(df['BOND']).all()

out = df[df['BOND'].isin(s[s].index)]

型
输出量：

BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL

型
如果出于某种原因，您可以使用不同的策略多次获得相同的STATUS，您可以使用set比较（此处为每组）：

s1 = df[df['STRATEGY'].eq('SELL')].groupby('BOND')['STATUS'].agg(set)
s2 = df.groupby('BOND')['STATUS'].agg(set)

s = s1.eq(s2)
out = df[df['BOND'].isin(s[s].index)]

型
或者将每个组与列中的整体可能策略进行比较：

s1 = df[df['STRATEGY'].eq('SELL')].groupby('BOND')['STATUS'].agg(set)
s2 = set(df['STATUS'])

s = s1.eq(s2)
out = df[df['BOND'].isin(s[s].index)]

型
备选输入：

BOND        STATUS STRATEGY
0    A        SENIOR      BUY
1    A  SUBORDINATED     SELL
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL
4    C        SENIOR      BUY
5    C        SENIOR     SELL
6    C  SUBORDINATED     SELL
7    D  SUBORDINATED     SELL

型
输出（每组）：

BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL
4    C        SENIOR      BUY # C was selected as there is at
5    C        SENIOR     SELL # least one SELL per STATUS
6    C  SUBORDINATED     SELL
7    D  SUBORDINATED     SELL # D is included as there is no SENIOR in the group

型
产出（总体）：
这里D被省略，因为没有匹配的'SENIOR'

BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL
4    C        SENIOR      BUY
5    C        SENIOR     SELL
6    C  SUBORDINATED     SELL

型

如果需要仅使用SELL策略测试所有BOND，用途：

out = df[~df['BOND'].isin(df.loc[df['STRATEGY'].ne('SELL'), 'BOND'])]
print (out)
  BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL

字符串
如果还需要测试BOND是否具有所有可能的STATUS值，请添加：

uniq = set(df['STATUS'])

out = out[out.groupby('BOND')['STATUS'].transform('nunique').eq(len(uniq))]
print (out)
  BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL

型
不同的输出与更改的数据：

print (df)
  BOND        STATUS STRATEGY
0    A        SENIOR      BUY
1    A  SUBORDINATED     SELL
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL
4    C        SENIOR      BUY
5    C        SENIOR     SELL
6    C  SUBORDINATED     SELL
7    D        SENIOR     SELL
8    D        SENIOR     SELL

out = df[~df['BOND'].isin(df.loc[df['STRATEGY'].ne('SELL'), 'BOND'])]
print (out)
  BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL
7    D        SENIOR     SELL
8    D        SENIOR     SELL
uniq = set(df['STATUS'])

out = out[out.groupby('BOND')['STATUS'].transform('nunique').eq(len(uniq))]
print (out)
  BOND        STATUS STRATEGY
2    B        SENIOR     SELL
3    B  SUBORDINATED     SELL

型