小编ron*_*000的帖子

我在Python3中编写了一个基本脚本来计算Collat​​z猜想.它采用正整数作为输入,并返回步骤的数字,直到序列下降到1.

我的脚本适用于任何小于2万亿的整数输入,但高于此阈值时输出太小.

举个例子,这里有一些输入,我的脚本的输出和实际的正确输出:

Integer Input          Script Output     Correct Output
   989,345,275,647        1,348             1,348 
 1,122,382,791,663        1,356             1,356 
 1,444,338,092,271        1,408             1,408 
 1,899,148,184,679        1,411             1,411 
 2,081,751,768,559          385             1,437 
 2,775,669,024,745          388             1,440 
 3,700,892,032,993          391             1,443 
 3,743,559,068,799          497             1,549 `

正确的输出值基于以下链接:http://www.ericr.nl/wondrous/delrecs.html

对于2万亿以上的输入,我的脚本输出总是比正确的输出少1,052,但我不知道是什么导致了这个.

谁能解释什么是错的,以及如何更新/修复脚本以使其适用于所有输入？我认为Python能够毫无问题地接受任意大数字...

谢谢!

# Python Code for the Collatz Conjecture
# Rules: Take any integer 'n' and assess:
# If integer is even, divide by 2 (n/2)
# If integer is odd, multiply by 3 and add 1 (3n+1)
# …

我正在尝试使用rmsle作为评估指标在 Python 中训练 lightgbm ML 模型，但当我尝试包含提前停止时遇到问题。

这是我的代码：

import numpy as np
import pandas as pd
import lightgbm as lgb
from sklearn.model_selection import train_test_split

df_train = pd.read_csv('train_data.csv')
X_train = df_train.drop('target', axis=1)
y_train = np.log(df_train['target'])

sample_params = {
    'boosting_type': 'gbdt',
    'objective': 'regression',
    'random_state': 42,
    'metric': 'rmsle',
    'lambda_l1': 5,
    'lambda_l2': 5,
    'num_leaves': 5,
    'bagging_freq': 5,
    'max_depth': 5,
    'max_bin': 5,
    'min_child_samples': 5,
    'feature_fraction': 0.5,
    'bagging_fraction': 0.5,
    'learning_rate': 0.1,
}

X_train_tr, X_train_val, y_train_tr, y_train_val = train_test_split(X_train, y_train, test_size=0.2, random_state=42)

def train_lightgbm(X_train_tr, y_train_tr, X_train_val, y_train_val, …