Pandas-根据列值查找一行中的第一次出现

Question

我有以下数据框：

 Row    Bid_price   Bid_volume  Ask_price   Ask_volume
 2      2999.0      786.7      -500.0       1403.2
 3      3000.0      786.7      -499.9       1407.2
 4      2950.0      787.3      -250.1       1407.2
---------------------
 56     125.1       2691       36.9         3113.1
 57     125         2691.1     37           3133.1
---------------------
 117    41.4        3029.7     2999         3835.7
 118    40.05       3029.7     3000         3835.7
---------------------
 123    39.4        3129.7     NaN          NaN
 124    36.1        3129.7     NaN          NaN
 125    36          3134.7     NaN          NaN

我需要取第一对Bid_price and Bid_volume (2999.0 and 786.7)并与所有对进行比较Ask_price and Ask_volume。只要Bid_volume < Ask_volume AND Bid_price > Ask_price我跳到下一对Bid_price and Bid_volume并再次与所有对Ask_price and Ask_volume. Bid_Price减少，Bid_Volume增加，Ask_Price增加，Ask_Volume增加。Bid_Price, Bid_Volume具有相同的长度，但Ask_Price和Ask_Volume更短。

输出应该是第一个实例 where Bid_volume is > Ask_volume AND Bid_price < Ask_price，因此满足条件。这是在壳体线124为一对Bid_Price and Bid_Volume，其与匹配线56为一对Ask_Price and Ask_Volume。

所需的输出应该是：

Row      Bid_price    Bid_volume  
124      36.1         3129.7

Row      Ask_price    Ask_volume
56       36.9         3113.1

我的问题是我只能评估每一行的条件。这不返回任何内容：

BidAsk = BidAsk[(BidAsk["Bid_volume"] > BidAsk["Ask_volume"]) & (BidAsk["Bid_price"] < BidAsk["Ask_price"])]
BidAsk[["Bid_price","Bid_volume"]]

这在这里给出了一个回溯错误：

BidAsk = BidAsk.where((BidAsk["Bid_volume"] > BidAsk["Ask_volume"]) & (BidAsk["Bid_Price"] < BidAsk["Ask_Price"]))
BidAsk[["Bid_price", "Bid_volume"]]

非常感谢任何帮助。谢谢！

Answer 1

我希望我理解正确，这个脚本会找到Bid_price并且Bid_volume满足条件Bid_volume is > Ask_volume AND Bid_price < Ask_price。

如果我有这个数据框：

   Bid_price  Bid_volume  Ask_price  Ask_volume
0     2999.0       786.7     -500.0      1403.2
1     3000.0       786.7     -499.9      1407.2
2     2950.0       787.3     -250.1      1407.2
3     2500.0       792.8     -250.0      1593.2
4     2000.0       798.9     -200.1      1593.2
5     1400.0      2000.0     1200.0      1600.0
6       36.1      3129.7        NaN         NaN

然后：

import pandas as pd
from io import StringIO

txt = '''Bid_price   Bid_volume  Ask_price   Ask_volume
2999.0      786.7      -500.0       1403.2
3000.0      786.7      -499.9       1407.2
2950.0      787.3      -250.1       1407.2
2500.0      792.8      -250.0       1593.2
2000.0      798.9      -200.1       1593.2
1400.0     2000.0      1200.0       1600.0
  36.1     3129.7             '''

df = pd.read_fwf(StringIO(txt))

max_price = df.Ask_price.max()
max_volume = df.Ask_volume.max()

mask = pd.concat([df.Bid_price < max_price, df.Bid_volume > max_volume], axis=1).all(axis=1)

print( df.loc[mask, ['Bid_price', 'Bid_volume']].head(1) )

印刷：

   Bid_price  Bid_volume
6       36.1      3129.7

编辑（根据更新的问题）：

import pandas as pd
from io import StringIO

txt = ''' Row    Bid_price   Bid_volume  Ask_price   Ask_volume
 2      2999.0      786.7      -500.0       1403.2
 3      3000.0      786.7      -499.9       1407.2
 4      2950.0      787.3      -250.1       1407.2
 56     125.1       2691       36.9         3113.1
 57     125         2691.1     37           3133.1
 117    41.4        3029.7     2999         3835.7
 118    40.05       3029.7     3000         3835.7
 123    39.4        3129.7     NaN          NaN
 124    36.1        3129.7     NaN          NaN
 125    36          3134.7     NaN          NaN'''

df = pd.read_fwf(StringIO(txt))

def get_indexes(df):
    for idx1, bid_price, bid_volume in zip(df.index, df.Bid_price, df.Bid_volume):
        for idx2, ask_price, ask_volume in zip(df.index, df.Ask_price, df.Ask_volume):
            if bid_volume > ask_volume and bid_price < ask_price:
                return idx1, idx2, bid_price, bid_volume, ask_price, ask_volume

print(df)
print()

result = get_indexes(df)
if result:
    print('Bid Price   =', result[2])
    print('Bid Volume  =', result[3])
    print('Ask Price   =', result[4])
    print('Ask Volume  =', result[5])
    print('Index bid   =', result[0])
    print('Index ask   =', result[1])

印刷：

   Row  Bid_price  Bid_volume  Ask_price  Ask_volume
0    2    2999.00       786.7     -500.0      1403.2
1    3    3000.00       786.7     -499.9      1407.2
2    4    2950.00       787.3     -250.1      1407.2
3   56     125.10      2691.0       36.9      3113.1
4   57     125.00      2691.1       37.0      3133.1
5  117      41.40      3029.7     2999.0      3835.7
6  118      40.05      3029.7     3000.0      3835.7
7  123      39.40      3129.7        NaN         NaN
8  124      36.10      3129.7        NaN         NaN
9  125      36.00      3134.7        NaN         NaN

Bid Price   = 36.1
Bid Volume  = 3129.7
Ask Price   = 36.9
Ask Volume  = 3113.1
Index bid   = 8
Index ask   = 3