Question

我有一个JSON文件，想将其解析为pandas DataFrame。我希望具有以下结构：

ID |大事记网站| ...

游戏1 | “额定子弹游戏” | https://lichess.org/spTUcy1Z | ...

Game2 | “额定闪电战” | https://lichess.org/kh6FJkyS | ...

但是使用df = pd.json_normalize(all_rows)，df看起来像这样：

--- | Game1.Event | Game1.Site | ...

0 |额定子弹游戏| https://lichess.org/spTUcy1Z | ...

JSON如下所示：

{
  "Game1": {
    "Event": "Rated Bullet game",
    "Site": "https://lichess.org/spTUcy1Z",
    "Date": "2020.05.01",
    "Round": "-",
    "White": "AaravShah25",
    "Black": "daksh_badhwar",
    "Result": "0-1",
    "UTCDate": "2020.05.01",
    "UTCTime": "04:55:40",
    "WhiteElo": "1360",
    "BlackElo": "1342",
    "WhiteRatingDiff": "-10",
    "BlackRatingDiff": "+6",
    "ECO": "C56",
    "Opening": "Italian Game: Scotch Gambit, Nakhmanson Gambit",
    "TimeControl": "60+0",
    "Termination": "Normal",
    "moves": "1. e4 { [%clk 0:01:00] } e5 { [%clk 0:01:00] } 2. d4 { [%clk 0:00:59] } Nc6 { [%clk 0:01:00] } 3. Nf3 { [%clk 0:00:58] } exd4 { [%clk 0:00:59] } 4. Bc4 { [%clk 0:00:58] } Nf6 { [%clk 0:00:56] } 5. O-O { [%clk 0:00:58] } Nxe4 { [%clk 0:00:54] } 6. Nc3 { [%clk 0:00:58] } Bc5 { [%clk 0:00:53] } 7. Re1 { [%clk 0:00:58] } O-O { [%clk 0:00:52] } 8. h3 { [%clk 0:00:56] } dxc3 { [%clk 0:00:50] } 0-1"
  },
  "Game2": {
    "Event": "Rated Blitz game",
    "Site": "https://lichess.org/kh6FJkyS",
    "Date": "2020.05.01",
    "Round": "-",
    "White": "Quggai",
    "Black": "vasiukov",
    "Result": "1-0",
    "UTCDate": "2020.05.01",
    "UTCTime": "07:41:06",
    "WhiteElo": "2292",
    "BlackElo": "2210",
    "WhiteRatingDiff": "+5",
    "BlackRatingDiff": "-4",
    "ECO": "C56",
    "Opening": "Italian Game: Scotch Gambit, Nakhmanson Gambit",
    "TimeControl": "180+0",
    "Termination": "Normal",
    "moves": "1. e4 { [%clk 0:03:00] } e5 { [%clk 0:03:00] } 2. Bc4 { [%clk 0:02:59] } Nf6 { [%clk 0:02:58] } 3. d4 { [%clk 0:02:58] } Nc6 { [%clk 0:02:51] } 4. Nf3 { [%clk 0:02:56] } exd4 { [%clk 0:02:48] } 5. O-O { [%clk 0:02:55] } Nxe4 { [%clk 0:02:46] } 6. Nc3 { [%clk 0:02:54] } dxc3 { [%clk 0:02:43] } 7. Bxf7+ { [%clk 0:02:48] } Kxf7 { [%clk 0:02:41] } 8. Qd5+ { [%clk 0:02:48] } Ke8 { [%clk 0:02:16] } 9. Re1 { [%clk 0:02:47] } Ne7 { [%clk 0:02:13] } 10. Rxe4 { [%clk 0:02:45] } c6 { [%clk 0:02:08] } 11. Qd6 { [%clk 0:02:30] } h6 { [%clk 0:01:32] } 12. Qg6# { [%clk 0:02:18] } 1-0"
  },
...

谢谢！

Answer 1

此JSON的结构看起来很容易使用默认的dataframe构造函数将其转换为dataframe：

df = pd.DataFrame(all_rows)

您还可以使用.transform (or just .T)函数来更改数据方向：

df = pd.DataFrame(all_rows).T

如果此JSON中有很多嵌套对象，则可以使用json_normalize从字典值创建数据框，然后设置正确的索引：

df = pd.json_normalize(list(all_rows.values()))
df.index = all_rows.keys()

熊猫json_normalize仅生成一大行

1 个答案: