Pandas 将多索引数据框转换为所有索引都在列中的简单数据框

我为这张照片道歉，但我不知道如何重现这张照片，因为我得到了这些数据

我只想将其转换为一个简单的数据帧，其中有索引列

time

，

lon

，

lat

，以及相应行中的值，如下所示：

| time | lat | lon | data |

我尝试过这样做。reset_index（），但是

时间轴仍然是横轴，而不是向下。如何“分解”所有索引值以获得一个简单的数据帧，其中所有索引都在列中

编辑：

用于复制的测试数据字典：

{Timestamp('2001-01-01 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -1.68,
  (50.18000030517578, -4.9200439453125): -1.88,
  (50.18000030517578, -4.219970703125): -2.08},
 Timestamp('2001-01-02 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -1.95,
  (50.18000030517578, -4.9200439453125): -2.25,
  (50.18000030517578, -4.219970703125): -2.55},
 Timestamp('2001-01-03 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -0.76,
  (50.18000030517578, -4.9200439453125): -0.91,
  (50.18000030517578, -4.219970703125): -1.06},
 Timestamp('2001-01-04 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -2.9,
  (50.18000030517578, -4.9200439453125): -3.01,
  (50.18000030517578, -4.219970703125): -3.11},
 Timestamp('2001-01-05 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -2.06,
  (50.18000030517578, -4.9200439453125): -2.29,
  (50.18000030517578, -4.219970703125): -2.52}}

选中此项：

import pandas as pd
from pandas import Timestamp


d = {Timestamp('2001-01-01 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -1.68,
  (50.18000030517578, -4.9200439453125): -1.88,
  (50.18000030517578, -4.219970703125): -2.08},
 Timestamp('2001-01-02 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -1.95,
  (50.18000030517578, -4.9200439453125): -2.25,
  (50.18000030517578, -4.219970703125): -2.55},
 Timestamp('2001-01-03 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -0.76,
  (50.18000030517578, -4.9200439453125): -0.91,
  (50.18000030517578, -4.219970703125): -1.06},
 Timestamp('2001-01-04 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -2.9,
  (50.18000030517578, -4.9200439453125): -3.01,
  (50.18000030517578, -4.219970703125): -3.11},
 Timestamp('2001-01-05 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -2.06,
  (50.18000030517578, -4.9200439453125): -2.29,
  (50.18000030517578, -4.219970703125): -2.52}}


df = pd.DataFrame(d)
df = df.stack().to_frame().reset_index()
df.columns = ['lat', 'lon', 'time', 'data']

输出：

      lat       lon       time  data
0   50.18 -5.619995 2001-01-01 -1.68
1   50.18 -5.619995 2001-01-02 -1.95
2   50.18 -5.619995 2001-01-03 -0.76
3   50.18 -5.619995 2001-01-04 -2.90
4   50.18 -5.619995 2001-01-05 -2.06
5   50.18 -4.920044 2001-01-01 -1.88
6   50.18 -4.920044 2001-01-02 -2.25
7   50.18 -4.920044 2001-01-03 -0.91
8   50.18 -4.920044 2001-01-04 -3.01
9   50.18 -4.920044 2001-01-05 -2.29
10  50.18 -4.219971 2001-01-01 -2.08
11  50.18 -4.219971 2001-01-02 -2.55
12  50.18 -4.219971 2001-01-03 -1.06
13  50.18 -4.219971 2001-01-04 -3.11
14  50.18 -4.219971 2001-01-05 -2.52

选中此项：

import pandas as pd
from pandas import Timestamp


d = {Timestamp('2001-01-01 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -1.68,
  (50.18000030517578, -4.9200439453125): -1.88,
  (50.18000030517578, -4.219970703125): -2.08},
 Timestamp('2001-01-02 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -1.95,
  (50.18000030517578, -4.9200439453125): -2.25,
  (50.18000030517578, -4.219970703125): -2.55},
 Timestamp('2001-01-03 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -0.76,
  (50.18000030517578, -4.9200439453125): -0.91,
  (50.18000030517578, -4.219970703125): -1.06},
 Timestamp('2001-01-04 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -2.9,
  (50.18000030517578, -4.9200439453125): -3.01,
  (50.18000030517578, -4.219970703125): -3.11},
 Timestamp('2001-01-05 00:00:00'): {(50.18000030517578,
   -5.6199951171875): -2.06,
  (50.18000030517578, -4.9200439453125): -2.29,
  (50.18000030517578, -4.219970703125): -2.52}}


df = pd.DataFrame(d)
df = df.stack().to_frame().reset_index()
df.columns = ['lat', 'lon', 'time', 'data']

输出：

      lat       lon       time  data
0   50.18 -5.619995 2001-01-01 -1.68
1   50.18 -5.619995 2001-01-02 -1.95
2   50.18 -5.619995 2001-01-03 -0.76
3   50.18 -5.619995 2001-01-04 -2.90
4   50.18 -5.619995 2001-01-05 -2.06
5   50.18 -4.920044 2001-01-01 -1.88
6   50.18 -4.920044 2001-01-02 -2.25
7   50.18 -4.920044 2001-01-03 -0.91
8   50.18 -4.920044 2001-01-04 -3.01
9   50.18 -4.920044 2001-01-05 -2.29
10  50.18 -4.219971 2001-01-01 -2.08
11  50.18 -4.219971 2001-01-02 -2.55
12  50.18 -4.219971 2001-01-03 -1.06
13  50.18 -4.219971 2001-01-04 -3.11
14  50.18 -4.219971 2001-01-05 -2.52

打印
df.head（5）.to_dict（）
的结果，从这里我可以
pd.DataFrame（…）
编辑-尽管我无法确定如何保存索引名（to_dict（）自动删除它们（？）-感谢您提供的帮助。我添加了解决方案。打印
df.head（5）的结果。从这里我可以
pd.DataFrame（）to_dict（）
（…）
Edited-虽然我无法确定如何保留索引名（to_dict（）自动删除它们（？）-感谢您提供的帮助。我添加了解决方案建议。
.stack（）
似乎是我想要的平展行索引！谢谢。有一件事需要改进：列名称处理，它应该更自动地完成；-）
.stack（）
似乎是我想要的平展行索引！谢谢。有一件事需要改进：列名处理，它应该更自动地完成；-）