In my dataframe I want to sum certain rows in a column and output them in a new column ‘UE_more_days’

is

    ATEXT   BEGUZ_UE   UE_more_days
0             11.00            0.0
1     CT      23.00            Nan
2     RT      33.00           46.0
3             15.00            0.0
3             15.00            0.0
4             12.75            0.0
5             19.75            0.0
6             14.75            0.0
7     CT      23.00           29.5
8     CT      24.00           46.0
9     CT      24.00           48.0
10    RT      33.00           48.0
11            15.00            0.0
12
etc

should be

    ATEXT   BEGUZ_UE   UE_more_days
0
1     CT      23.00
2     RT      33.00           56.0
3             15.00
4             12.75
5             19.75
6             14.75
7     CT      23.00
8     CT      24.00
9     CT      24.00
10    RT      33.00          104.0
11            15.00
12
etc

my last try

bedd2 = [(df['ATEXT'] != ''),]
result2 = [(df.iloc[0:]['BEGUZ_UE'].astype(float).reset_index(drop=True) +
df.iloc[1:]['BEGUZ_UE'].astype(float)).round(decimals=2).shift(1)]
df['min_UE_mehr_Tage'] = np.select(bedd2, result2)

How can I sum rows from a column based on a condition and output them in a new column?