从列表中查找每个帐户的平均值和最大值

Question

所以我得到了 'list' 类型的数据（这只是它的一小部分来说明这一点）存储在一个名为 sorted_trans 的变量中（我已经预先对它进行了排序将 accountId 组合在一起）：

Transaction(transactionId='T000664', accountId='A1', transactionDay=21, category='GG', transactionAmount=364.69), Transaction(transactionId='T000776', accountId='A1', transactionDay=24, category='GG', transactionAmount=329.63), Transaction(transactionId='T000313', accountId='A10', transactionDay=8, category='AA', transactionAmount=960.56), Transaction(transactionId='T000472', accountId='A10', transactionDay=12, category='AA', transactionAmount=707.74), Transaction(transactionId='T000596', accountId='A10', transactionDay=18, category='AA', transactionAmount=156.77), Transaction(transactionId='T000730', accountId='A10', transactionDay=23, category='AA', transactionAmount=577.76)

我的任务是将过去 5 天的数据按 'accountId' 排序（我正在使用滚动 window）并计算：总计 'transactionAmount' 每 'accountId'，平均 'transactionAmount' 每 'accountId'，最大 'transactionAmount' 每 'accountId'

return

为了捕获过去五天（不包括当天）的数据，我使用的是普通的 for 循环

for i in range(day-window_size, day):

我怀疑我需要使用某种理解来分组和计算正确的值。输出应包含每个账户 ID 每天一行，每一行应包含每个计算的统计信息，例如：Output examle。除了平均值和最大值，我已经设法得到了所有东西。

例如，我使用以下代码计算了每个类别 AA（其他类别相同）的总价值：

trans_AA = []
trans_AA.append([(x.accountId, x.transactionAmount) for x in sorted_trans  if x.category == "AA"])

计算每个类别总交易金额的通用方法

def totals_per_cat(transactions):
    
    for accountId, transactionAmount in chain.from_iterable(transactions):
                 d[accountId] += transactionAmount
    trans_res = list(d.values())
    return trans_res

如何计算过去 5 天每个 accountId 的平均值和最大值？

Answer 1

这有望帮助您完成大部分工作：

from collections import defaultdict
from dataclasses import dataclass, fields
from csv import DictReader
from statistics import mean
from typing import Dict, List



@dataclass
class Transaction:
    transactionId: str
    accountId: str
    transactionDay: int
    category: str
    transactionAmount: float 


def load_transactions(file_path: str) -> List[Transaction]:
    with open(file_path) as csv_file:
        return [
            Transaction(**{
                field.name: field.type(row[field.name])
                for field in fields(Transaction)
            }) for row in DictReader(csv_file)
        ]

transactions = load_transactions('transactions.txt')
min_day = min(xact.transactionDay for xact in transactions)
max_day = max(xact.transactionDay for xact in transactions)
window_size = 5

xact_by_acct: Dict[str, List[Transaction]] = defaultdict(list)
for xact in transactions:
    xact_by_acct[xact.accountId].append(xact)

for day in range(min_day + window_size, max_day + 1):
    for acct, xacts in xact_by_acct.items():
        amts = [
            xact.transactionAmount for xact in xacts 
            if xact.transactionDay in range(day - window_size, day)
        ]  # all amounts within the window
        try:
            print(
                f"Day: {day}\tAccount: {acct}\t"
                f"Max: {max(amts)}\tAverage: {mean(amts)}"
            )
        except ValueError:
            # No transactions in the window for this account.
            pass

从列表中查找每个帐户的平均值和最大值

Finding average and max value per account from a list

python

transactions

max

list-comprehension

average