我正在嘗試捕獲列表格式的數據框/熊貓中的元素。如果字符串存在，下面會捕獲整個列表，我如何僅按行捕獲特定字符串的元素而忽略其余部分？這是我嘗試過的...l1 = [1,2,3,4,5,6]l2 = ['hello world \n my world','world is a great place \n we live in it','planet earth',np.NaN,'\n save the water','']df = pd.DataFrame(list(zip(l1,l2)), columns=['id','sentence'])df['sentence_split'] = df['sentence'].str.split('\n')print(df)這段代碼的結果：df[df.sentence_split.str.join(' ').str.contains('world', na=False)] # does the trick but still not exactly what I am looking for. id sentence sentence_split1 hello world \n my world [hello world , my world]2 world is a great place \n we live in it [world is a great place , we live in it]但尋找：id sentence sentence_split1 hello world \n my world hello world; my world2 world is a great place \n we live in it world is a great place
1 回答
UYOU
TA貢獻1878條經驗 獲得超4個贊
您正在尋找在一系列列表中搜索字符串。一種方法是：
# Drop NaN rows
df = df.dropna(subset=["sentence_split"])
應用僅保留您要查找的列表中的元素的函數
# Apply this lamda function
df["sentence_split"] = df["sentence_split"].apply(lambda x: [i for i in x if "world" in i])
id sentence sentence_split
0 1 hello world \n my world [hello world , my world]
1 2 world is a great place \n we live in it [world is a great place ]
2 3 planet earth []
4 5 \n save the water []
5 6 []
添加回答
舉報
0/150
提交
取消