要使用纯python写

如果您需要使用纯Python编写代码来完成《大数据分析方法》实验报告中的任务，以下是一些关键步骤和技术要点：

### 实验一：关联规则算法实验
#### 目标
使用Apriori算法发现蔬菜价格的相关性。

#### 步骤
1. **数据预处理**
- 读取CSV文件，处理缺失值。
- 将数据转换为适合Apriori算法的格式。

2. **实现Apriori算法**
- 计算频繁项集。
- 生成关联规则。

3. **结果分析**
- 解释发现的关联规则，分析蔬菜价格的同涨、同跌或涨跌不同步现象。

#### 示例代码

import csv
from collections import defaultdict

def load_data(file_path):
    transactions = []
    with open(file_path, 'r') as file:
        reader = csv.DictReader(file)
        for row in reader:
            transaction = [row['蔬菜名'], row['价格']]
            if all(transaction):  # 过滤掉有空值的行
                transactions.append(transaction)
    return transactions

def apriori(transactions, min_support=0.5):
    def generate_candidates(itemsets, k):
        candidates = set()
        for itemset in itemsets:
            for other_itemset in itemsets:
                candidate = itemset.union(other_itemset)
                if len(candidate) == k and candidate not in candidates:
                    candidates.add(frozenset(candidate))
        return candidates

    def prune(candidates, prev_freq_itemsets, k):
        pruned_candidates = set()
        for candidate in candidates:
            subsets = [frozenset(x) for x in combinations(candidate, k-1)]
            if all(subset in prev_freq_itemsets for subset in subsets):
                pruned_candidates.add(candidate)
        return pruned_candidates

    from itertools import combinations
    C1 = {frozenset([item]) for transaction in transactions for item in transaction}
    L1 = {item: sum(1 for t in transactions if item.issubset(t)) / len(transactions) for item in C1}
    Lk = {item for item, support in L1.items() if support >= min_support}

    freq_itemsets = list(Lk)
    k = 2
    while True:
        Ck = generate_candidates(Lk, k)
        Ck = prune(Ck, Lk, k)
        Lk = {item: sum(1 for t in transactions if item.issubset(t)) / len(transactions) for item in Ck}
        Lk = {item for item, support in Lk.items() if support >= min_support}
        if not Lk:
            break
        freq_itemsets.extend(Lk)
        k += 1

    return freq_itemsets

def generate_rules(freq_itemsets, min_confidence=0.7):
    rules = []
    for itemset in freq_itemsets:
        for i in range(1, len(itemset)):
            for antecedent in combinations(itemset, i):
                antecedent = frozenset(antecedent)
                consequent = itemset.difference(antecedent)
                confidence = support[frozenset(itemset)] / support[antecedent]
                if confidence >= min_confidence:
                    rules.append((antecedent, consequent, confidence))
    return rules

# 加载数据
transactions = load_data('vegetable_prices.csv')

# 计算频繁项集
freq_itemsets = apriori(transactions)

# 生成关联规则
rules = generate_rules(freq_itemsets)

# 打印结果
for rule in rules:
    print(f"Rule: {rule[0]} -> {rule[1]}, Confidence: {rule[2]:.2f}")

### 其他实验
对于其他实验（如回归算法、分类算法、聚类算法），您可以参照类似的步骤：
1. **数据预处理**：读取数据，处理缺失值，标准化或归一化数据。
2. **算法实现**：选择合适的算法并实现。
3. **结果分析**：解释模型的性能和结果。

如果您需要更详细的代码示例或其他帮助，请告诉我具体的需求。