python字典类型的维度_迭代python字典中的一个维度

i created a dictionary on 2 dimensions

myaddresses['john','smith'] = "address 1"

myaddresses['john','doe'] = "address 2"

How can i iterate over one dimension in the fashion

for key in myaddresses.keys('john'):

解决方案

Bad news: you can't (not directly at least). What you did was not a "2 dimensions" dict, but a dict with tuples (string pairs in your case) as keys, and only the hash value of the key is used (as usually with hashtables). What you want requires a sequential lookup, ie:

for key, val in my_dict.items():

# no garantee we have string pair as key here

try:

firstname, lastname = key

except ValueError:

# not a pair...

continue

# this would require another try/except block since

# equality test on different types can raise anything

# but let's pretend it's ok :-/

if firstname == "john":

do_something_with(key, val)

Needless to say that it kind of defeat the whole point of using a dict. Err... what about using a proper relational DB instead ?

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
ALS(Alternating Least Squares)算法是一种常用的协同过滤推荐算法。下面是一个简单的Python实现过程: ```python import numpy as np # 定义ALS类 class ALS: def __init__(self, ratings, K, alpha, lam, max_iter): """ :param ratings: 用户评分矩阵,字典类型,键为用户ID,值为字典类型,键为物品ID,值为评分 :param K: 隐向量维度 :param alpha: 学习率 :param lam: 正则化参数 :param max_iter: 最大迭代次数 """ self.ratings = ratings self.K = K self.alpha = alpha self.lam = lam self.max_iter = max_iter # 初始化用户向量和物品向量 self.user_vectors, self.item_vectors = self.init_vectors() def init_vectors(self): """ 初始化用户向量和物品向量 """ user_vectors = {} item_vectors = {} for user_id in self.ratings.keys(): user_vectors[user_id] = np.random.normal(size=self.K) for item_id in self.ratings[self.ratings.keys()[0]].keys(): item_vectors[item_id] = np.random.normal(size=self.K) return user_vectors, item_vectors def train(self): """ 训练模型 """ for step in range(self.max_iter): for user_id in self.ratings.keys(): for item_id in self.ratings[user_id].keys(): # 预测评分 rating_hat = np.dot(self.user_vectors[user_id], self.item_vectors[item_id]) # 计算误差 error = self.ratings[user_id][item_id] - rating_hat # 更新用户向量和物品向量 self.user_vectors[user_id] += self.alpha * (error * self.item_vectors[item_id] - self.lam * self.user_vectors[user_id]) self.item_vectors[item_id] += self.alpha * (error * self.user_vectors[user_id] - self.lam * self.item_vectors[item_id]) def predict(self, user_id, item_id): """ 预测用户对物品的评分 """ rating_hat = np.dot(self.user_vectors[user_id], self.item_vectors[item_id]) return rating_hat ``` 以上是一个简单的Python实现过程,更详细的注释请参考代码。如果你想了解更多关于ALS算法的知识,可以参考推荐系统领域的相关书籍和论文。

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值