如何从Transaction类实例创建指定列的Pandas DataFrame？

首先先修正你代码里的几个问题：

1. 类定义名称是Transactions，但实例化时用了Transaction，名称不统一，需改成一致（比如统一为Transaction）
2. add_money类方法的参数用了self，会和实例的self冲突，建议改为transaction
3. 类属性amount和实例属性amount重名，容易混淆，改为total_amount更清晰

修正后的类代码：

class Transaction:
    num_of_transactions = 0
    total_amount = 0
    # 新增类属性存储所有交易实例，方便后续批量处理
    all_transactions = []

    def __init__(self, date, concept, amount):
        self.date = date
        self.concept = concept
        self.amount = amount
        Transaction.add_transaction()
        Transaction.add_money(self)
        Transaction.all_transactions.append(self)

    @classmethod
    def number_of_transactions(cls):
        return cls.num_of_transactions

    @classmethod
    def add_transaction(cls):
        cls.num_of_transactions += 1

    @classmethod
    def total_amount_of_money(cls):
        return cls.total_amount

    @classmethod
    def add_money(cls, transaction):
        cls.total_amount += transaction.amount

# 实例化交易对象
t1 = Transaction("20221128", "C1", 14)
t2 = Transaction("20221129", "C2", 30)
t3 = Transaction("20221130", "C3", 14)

接下来介绍两种创建符合要求的DataFrame的方法：

方法一：手动收集实例属性生成DataFrame #

直接提取每个实例的date、concept、amount属性，构造字典传入pd.DataFrame：

import pandas as pd

data = [
    {"Date": t1.date, "Concept": t1.concept, "Amount": t1.amount},
    {"Date": t2.date, "Concept": t2.concept, "Amount": t2.amount},
    {"Date": t3.date, "Concept": t3.concept, "Amount": t3.amount}
]

df = pd.DataFrame(data)
print(df)

方法二：通过类的统一列表自动生成DataFrame #

利用类里新增的all_transactions属性，批量提取所有实例的属性，适合后续新增交易实例的场景：

import pandas as pd

data = [
    {"Date": trans.date, "Concept": trans.concept, "Amount": trans.amount}
    for trans in Transaction.all_transactions
]

df = pd.DataFrame(data)
print(df)

两种方法最终都会生成包含Date、Concept、Amount三列的DataFrame，输出结果如下：

Date Concept  Amount
0  20221128      C1      14
1  20221129      C2      30
2  20221130      C3      14

内容的提问来源于stack exchange，提问作者Adrià Martínez