主页 / 问题 / 1079155
Accepted
Alpensin
Asked:2020-02-06 20:03:25 +0000 UTC2020-02-06 20:03:25 +0000 UTC

按源 DataFrame 转换为列表列表

  • 772

请告诉我这是如何以 Pandas 的风格精美完成的。到目前为止，我只能在可怕的周期的帮助下想象。初始数据：

import pandas as pd
import numpy as np
from datetime import datetime
data= {'ID': {0: 308, 1: 308, 2: 309, 3: 309},
 'LON': {0: 37.409936, 1: 37.409936, 2: 37.259936, 3: 37.259936},
 'LAT': {0: 55.967077, 1: 55.967077, 2: 55.997077, 3: 55.997077},
 'PERIOD_TIME': {0: datetime.strptime('2020-02-05 00:00:00', '%Y-%m-%d %H:%M:%S'),
  1: datetime.strptime('2020-02-05 01:00:00', '%Y-%m-%d %H:%M:%S'),
  2: datetime.strptime('2020-02-05 00:00:00', '%Y-%m-%d %H:%M:%S'),
  3: datetime.strptime('2020-02-05 01:00:00', '%Y-%m-%d %H:%M:%S')},
 'WEIGHT': {0: 0.1, 1: 0.9, 2: 0.4, 3: 0.5}}
df = pd.DataFrame(data)

df 看起来像这样：

ID  LON LAT PERIOD_TIME WEIGHT
0   308 37.409936   55.967077   2020-02-05 00:00:00 0.1
1   308 37.409936   55.967077   2020-02-05 01:00:00 0.9
2   309 37.259936   55.997077   2020-02-05 00:00:00 0.4
3   309 37.259936   55.997077   2020-02-05 01:00:00 0.5

需要创建一个单独的唯一 PERIOD_TIME 列表：

['2020-02-05 00:00:00', '2020-02-05 01:00:00']

以及来自数据的列表 [LON, LAT, WEIGHT]：

[[[37.409936, 55.967077, 0.1],[37.259936, 55.997077, 0.4]],
[[37.409936, 55.967077, 0.9],[37.259936, 55.997077, 0.5]]]
python

2 个回答

  1. Best Answer

    可以这样做：

    In [50]: res = df.groupby("PERIOD_TIME")["LON","LAT","WEIGHT"].apply(lambda x: x.values.tolist())

In [51]: res
Out[51]:
PERIOD_TIME
2020-02-05 00:00:00    [[37.409936, 55.967077, 0.1], [37.25...
2020-02-05 01:00:00    [[37.409936, 55.967077, 0.9], [37.25...
dtype: object

In [52]: res.iloc[0]
Out[52]: [[37.409936, 55.967077, 0.1], [37.259936, 55.997077, 0.4]]

    或像这样：

    res = (df
       .assign(PERIOD_TIME=df["PERIOD_TIME"].astype(str))
       .groupby("PERIOD_TIME")
       ["LON","LAT","WEIGHT"]
       .apply(lambda x: x.values.tolist())
       .to_dict())

In [58]: res
Out[58]:
{'2020-02-05 00:00:00': [[37.409936, 55.967077, 0.1],
  [37.259936, 55.997077, 0.4]],
 '2020-02-05 01:00:00': [[37.409936, 55.967077, 0.9],
  [37.259936, 55.997077, 0.5]]}
    • 3

  2. 对于唯一的列表PERIOD_TIME：

    In[5]: lst = df["PERIOD_TIME"].drop_duplicates().to_list()
In[6]: lst

    Out[6]: [Timestamp('2020-02-05 00:00:00'), Timestamp('2020-02-05 01:00:00')]

    In[7]: [ts.strftime("%Y-%m-%d %H-%M-%S") for ts in lst]

    Out[7]: ['2020-02-05 00-00-00', '2020-02-05 01-00-00']

    对于来自 data 的列表列表[LON, LAT, WEIGHT]：

    In[8]: df1 = df[["LON", "LAT", "WEIGHT"]].T
In[9]: df1

    Out[9]: 
                0          1          2          3
LON     37.409936  37.409936  37.259936  37.259936
LAT     55.967077  55.967077  55.997077  55.997077
WEIGHT   0.100000   0.900000   0.400000   0.500000

    In[10]: [df1[i].to_list() for i in range(4)] 

    Out[10]: 
[[37.409936, 55.967077, 0.1],
 [37.409936, 55.967077, 0.9],
 [37.259936, 55.997077, 0.4],
 [37.259936, 55.997077, 0.5]]
    • 2

相关问题