Question

我将数据保存在pandas.groupby对象中，并尝试根据标题为“金额”的条件栏中的条件遍历各个组。但是，我收到的错误消息正在尝试将“参考”代码从字符串转换为浮点数，但是我不确定该指令的位置。

for data in row:
    if float(data) in ['Amount'] > 0:
        {'buy_currency' : ['Currency'],
          'buy_quantity' : ['Amount'],
        'order_id' : str(data)['Reference']}

我认为我误会了第二行：if float(data) in ['Amount'] > 0

我只想将“金额”字段转换为浮点数，其他字段保留字符串即可。

我感谢任何在正确方向上的指导或推动！

import pandas as pd
import numpy as np

df2 = pd.read_excel('sample.xlsx')
mask = df2.groupby(['Reference'])
groups = mask.groups

for ref, trades in mask:
    for index, row in trades.iterrows():
       for data in row:
                if float(data) in ['Amount'] > 0:
                    {'buy_currency' : ['Currency'],
                     'buy_quantity' : ['Amount'],
                     'order_id' : str(data)['Reference']}
                else:
                    {'sell_currency' : ['Currency'],
                     'buy_quantity' : ['Amount'],
                     'order_id': str(data)['Reference']}

数据集包含以下样本： Dataset Sample

我收到的错误消息是：

ValueError: could not convert string to float: 'bbb8ee04-053c-4174-b5b1-281c10618d52'

Answer 1

我不确定您要做什么。您是否知道可以在熊猫中使用masks一次检查多个列的值？以下是一些可能对您有用的示例代码：

import pandas as pd
import numpy as np

df2 = pd.read_excel('sample.xlsx')
df2['Amount'] = df2['Amount'].astype(float)
df2['Reference'] = df2['Reference'].astype(str)

# no need to group
mask = df2['Amount'] > 0
df2['buy_currency'] = np.where(mask, df2['Currency'], np.nan)
df2['buy_quantity'] = np.where(mask, df2['Amount'], np.nan)
df2['order_id'] = np.where(mask, df2['Reference'], np.nan)

mask2 = df2['Amount'] < 0
df2['sell_currency'] = np.where(mask2, df2['Currency'], np.nan)
df2['sell_quantity'] = np.where(mask2, df2['Amount'], np.nan)
df2['order_id'] = np.where(mask2, df2['Reference'], np.nan)

ValueError（浮点型字符串），有条件的For循环

1 个答案: