根据以下行中的值连接列中的行

Question

我有一只像这样的熊猫：

df1 = pd.DataFrame({"strings":pd.Series(["a very ", "very boring text", " I cannot read it", "Hi everyone"," please go home ","or I will go ","now"]),
                "changetype":pd.Series([0,0,-1,0,1,1,1])})

每次行中的changetype == row + 1，我想要连接字符串。因此，最终的df应如下所示：

df2 = pd.DataFrame({"strings":pd.Series(["a very very boring text", " I cannot read it", "Hi everyone"," please go home or I will go now"]),
                "changetype":pd.Series([0,-1,0,1,])})

我是这样开始的，但不知道如何继续：

for row, rowplus in zip(df1.changetype, df1.changetype[1:]):
if row == rowplus:
    # concat rows here ...

Answer 1

使用汇总i128和Series的帮助first groupby：

join

<强>解释：

将shift d列与ne s = df1['changetype'].ne(df1['changetype'].shift()).cumsum() df3 = df1.groupby(s).agg({'strings':' '.join, 'changetype':'first'}).reset_index(drop=True) print (df3) strings changetype 0 a very very boring text 0 1 I cannot read it -1 2 Hi everyone 0 3 please go home or I will go now 1 进行比较，并为连续组!=添加cumsum：

Series