由买买提看人间百态

topics

全部话题 - 话题: bisection
1 2 下页 末页 (共2页)
i******s
发帖数: 301
1
来自主题: JobHunting版 - triplets 题怎么做比较高效?
这不就是抄interviewstreet上的么.
下面是我的python code,idea一样,处理重复有一点技巧,你可以自己想想。Time
complexity 应该是O(n^2),主要花在build smaller, greater array上,不过应该能
做到O(nlogn) on average case。我没仔细想过。
import sys
import bisect
arr_size = int(sys.stdin.readline())
arr = map(int, sys.stdin.readline().split())
smaller_than = [0 for i in range(arr_size)]
greater_than = [0 for i in range(arr_size)]
index_map = {}
for i in range(arr_size):
if arr[i] in index_map:
index_map[arr[i]].append(i)
else:
index_map[arr[... 阅读全帖
t********n
发帖数: 611
2
来自主题: Programming版 - 求教一个程序问题
自学编程遇到困难,哪位前辈帮忙看看:
问题是:要求写一个程序,计算如果每月还固定金额,一年把信用卡债务还清,每月需
要支付多少钱,前提是这一年都不再使用信用卡借债。要求用bisection
lower bound= outstanding balance/12
upper bound=( outstanding balance (1+monthly interest rate)**12)/12
以下是问题link:
http://ocw.mit.edu/courses/electrical-engineering-and-computer-
第3题不会做,我写的程序是这样的:
initialBalance= float (raw_input ('Enter the outstanding balance:'))
annualInterest= float (raw_input('Enter the annula interest rate'))
monthlyInterest= annualInterest/12.0
lower= initialBalance/12.0
upper= ... 阅读全帖
m****e
发帖数: 7
3
yes, thanks. I used bisection to "guess" the IV in matlab... but always
wondered if there exists closed form to get IV easily.
So is it usually done in such numerical way in the real world?
I am new in this field but I guess you may run into this problem quite a lot
in real world... and I wonder what's the most efficient way to figure out
the IV from option prices... besides bisection...?
Thanks!

Newton or bisection to find the solution. Since vanilla option price is a
monotonic function of imp
k*z
发帖数: 4704
4
来自主题: Statistics版 - 牛牛 Xie Liang关于 Erlang C的Model应用
http://saslist.com/blog/2012/07/13/sas-functions-for-computing-
http://listserv.uga.edu/cgi-bin/wa?A2=ind1210c&L=sas-l&F=&S=&P=
SAS functions for computing parameters in Erlang-C model
Call center management is both Arts and Sciences. While driving moral and
setting up strategies is more about Arts, staffing and servicing level
configuration based on call load is in the domain of Sciences.
The science part of call center management is based on Queueing Theory,
which studies "the Phenomena of sta... 阅读全帖
Y*******n
发帖数: 2296
5
来自主题: History版 - 大家谁看过蝴蝶君?
儿子时度度是从新疆找的维族小巴郎。
时佩普的一个堂弟,讨的是维族老婆,给联络的。
NY Times写时佩璞有特殊功能,能把鸡巴收起来。
‘Easily, smoothly, he pushes his testicles up into his body cavity. The
skin of the scrotal sack hangs slack, like curtains. The man now pushes his
penis between his legs, toward his back, bisecting the skin of the scrotum,
and squeezes his legs tightly together. The penis is hidden, while the skin
of the scrotum resembles the vaginal lips, beneath a triangle of pubic hair.
Pushed between the empty scrotal sac, the penis has als... 阅读全帖
Y*******n
发帖数: 2296
6
这是我几年来看到的最娱乐最搞笑的故事。
不知大家看过M. Butterfly这部电影没有。
汉语翻译蝴蝶夫人,蝴蝶君。
http://www.imdb.com/title/tt0107468/
这部电影基本上是写实的,讲述了时佩璞的传奇故事。
一下引百度百科
’时佩璞,男,是北京青年京剧团的编剧、旦角演员兼团部秘书。时佩璞毕业于云南大
学,主攻法语兼西班牙语,学生时代就喜欢京剧,曾与关肃霜合作演出。他后来拜小生
姜妙香为师,曾经在北大礼堂演出《奇双会》。
1964年认识当时驻北京的法国外交官Boursicot,对方以为他是女儿身, 猛烈追求他,
两人其后更发生性关系。在北京工作的布尔西科,一次在使馆举办的酒会期间遇上男扮
女装的花旦演员时佩璞,两人陷入爱河,但布尔西科一直都不知道时佩璞是男儿身,两
人密切交往后,时佩璞更告诉布尔西科,已经怀了他的儿子。
1969年,情报机构利用这关系,将Boursicot招为谍报人员。
Boursicot于1969到1972年派驻北京,
1977到1979年委派到蒙古乌兰巴托,他向中国提供超过500份秘密情报。 由于
Boursicot常驻中国之外,与... 阅读全帖
X******i
发帖数: 1384
7
来自主题: Military版 - 法国科学家
谢谢,我没干过数值计算的活,这次是赶鸭子上架。我大学是在法国读的,所以有些中
文术语不清楚,牛顿迭代是不是就是newton raphson?收敛应该是converge吧?
我的问题有点复杂,前面没有解释清楚。关键是逼近值一定要准,因为这只是一个单元
里的逼近值,然后输出到下一个单元作为输入。一共有几千个单元,所以如果不准的话
最后的误差会被放大。我最开始用的是brute forcing,一步一步向交点逼近,计算停止
在误差小于0.1%(举例),但是如果我的step取的不够小的话,可能最靠近交点的逼近
值都不能在0.1%以内,无法converge。但是step取的小的话,计算量又很大。
bisection我也用过,也不准尤其是在多个单元之后,误差会越来越大。
我还试过shooting method,其实差不多就是newton raphson了,两边取tangent line向
中间逼近,但是他的缺点是逼近点在不同的单元里有时在左侧,有时在右侧,不过倒是
有可能互相取消。但是横坐标的数值不是整数,给我后面的运算带来很多麻烦。
c*********d
发帖数: 9770
8
来自主题: Military版 - 中国“私募一哥”的覆灭
http://cn.nytimes.com/business/20160330/c30mag-hedgefund1/dual/
The Fall of China’s Hedge Fund King
By ALEX W. PALMER March 30, 2016
中国“私募一哥”的覆灭
ALEX W. PALMER 2016年3月30日
At 10:33 a.m. on Sunday, Nov. 1 of last year, the highway police in Ningbo,
an industrial city on China’s eastern coast, posted a seemingly innocuous
message to their official microblog. “Due to sudden traffic control,” the
message read, “all the entrances and exits of the G15 Expressway at
Hangzhou Bay Bridge have been closed.... 阅读全帖

发帖数: 1
9
来自主题: Military版 - 在课堂上讲“希腊伪史论”
你以为所谓newton bisection method从哪来的?
朱载堉会算2的根号下十二次方
那时候牛顿还是液体
c**i
发帖数: 6973
10
来自主题: USANews版 - Test Your Mathematics Skill
Do not peek at the answer at the lower right corner, which in the print of
report below is printed upside down. Here is hte question.
http://www.nytimes.com/imagepages/2010/04/15/us/15math_CA0.html
Note:
(a) parallelogram (n): "a quadrilateral with opposite sides parallel and
equal"
* quadrilateral (n): "a polygon of four sides"
(b) bisect (vt): "to divide into two usually equal parts"
l****z
发帖数: 29846
11
I see the G8 has a brilliant solution to the problems of the eurozone.
President Obama says it’s time for “growth and jobs”. Jolly good. That’s
the stuff. Let me show you how to create employment – the Brussels way.
Come with me through the streets of Athens, not far from Syntagma Square,
and your mind will reel with the horrified realisation that history is not a
one-way ratchet, that human progress is not guaranteed, and that a proud
country can be reduced – by years of torture and bullying – ... 阅读全帖
l****z
发帖数: 29846
12
来自主题: USANews版 - Keystone report出来了
By: Deena Winter
2/3/2014 10:58 AM
This article originally appeared on watchdog.org.
LINCOLN, Neb. — The U.S. State Department’s final environmental review of
the proposed Keystone XL oil pipeline mirrors earlier conclusions that the
pipeline wouldn’t significantly contribute to greenhouse gas emissions.
The report reiterated last year’s draft report conclusion that the pipeline
is unlikely to significantly impact the rate of extraction of oil sands or
the continued demand for heavy crude oil... 阅读全帖
i****x
发帖数: 17565
13
Officer Damon Badnell was responding to an officer involved in a foot
pursuit of a burglary suspect when he lost control of his vehicle on a
slippery road. The momentum wrapped the vehicle around the tree, nearly
bisecting the Ford Crown Victoria.
收到同伙请求正要去抓坏蛋时打滑,肯定不是highway否则一般会说,而且crash照片很
明显不在高速上或者高速旁边。所以速度最多也就是五十迈。
a***g
发帖数: 234
14
这个不就直接bisection么
没有其他条件么

(0
the
r*****e
发帖数: 7853
15
来自主题: JobHunting版 - 请教一道题
sorted, use bisection, O(logn)

to
r********g
发帖数: 144
16
来自主题: JobHunting版 - 求sqrt的binary算法,多谢
I thought that this method is called the bisection method.
http://en.wikipedia.org/wiki/Bisection_method
q**r
发帖数: 611
17
来自主题: JobHunting版 - 如何计算sqrt
Sqrt(x) = ?
Find a, b, s.t., a^2 < x < b^2.
Then Bisection between a & b.
w*****e
发帖数: 158
18
来自主题: JobHunting版 - 两道算法题
1) given the whole range of 6 digit numbers, design an algorithm which
prints out numbers which have the sum of the right 3 digits equal the left
3 digits? like 123600, 345435 etc.
2)design an algorithm to find square_root of a float? how do you make the
algorithm efficient?
第一题:除了每个数检验之外,还有什么更快得方法吗?
第二题:貌似可以用bisection method 或者 Newton method, Newton method 收敛的
更快
一些。有什么地方需要注意或者可以改进,使得算法更有效?
希望大家指正,先谢谢了。
n*****y
发帖数: 361
19
来自主题: JobHunting版 - Offer + 很多面经
因为赶着年底毕业,九月底才开始投简历。这个offer来的太快,小startup就是动作快
,从十月初联系我,到 offer, 就两周。那几个大公司的 on-campus interview还都没
开始。也算是 hot startup,但这里肯定没人知道的,移民版知道这个公司的更多些。
就不透露公司的名字和考题了,见谅。
HR 联系之后,先是组里的头直接电面,问了一个他们实践中的问题,我没想出答案,
但还是扯了扯。后来就在谈公司做什么,我有准备,问了很多问题。刚放下电话,HR问
我什么时候作 coding test, 可以马上把题发给我, 就是fill Java class, 实现某些
功能,一般给 2-3个小时。我想这么大工作量,不能拖,否则牵扯时间和精力,就说马
上作,决定不准备了,冒一定风险。结果一个小时做完发给他们,小impress了一下。
然后另一个 team lead 马上打电话二面,顺便考察一下是不是真是我自己写的程序。
这就是一天三面,两电一编程,然后就给 onsite了。面了组里的4个lead,HR 和
Founder。前两个基本都在问我的 research和 big ... 阅读全帖
n*****y
发帖数: 361
20
来自主题: JobHunting版 - Offer + 很多面经
因为赶着年底毕业,九月底才开始投简历。这个offer来的太快,小startup就是动作快
,从十月初联系我,到 offer, 就两周。那几个大公司的 on-campus interview还都没
开始。也算是 hot startup,但这里肯定没人知道的,移民版知道这个公司的更多些。
就不透露公司的名字和考题了,见谅。
HR 联系之后,先是组里的头直接电面,问了一个他们实践中的问题,我没想出答案,
但还是扯了扯。后来就在谈公司做什么,我有准备,问了很多问题。刚放下电话,HR问
我什么时候作 coding test, 可以马上把题发给我, 就是fill Java class, 实现某些
功能,一般给 2-3个小时。我想这么大工作量,不能拖,否则牵扯时间和精力,就说马
上作,决定不准备了,冒一定风险。结果一个小时做完发给他们,小impress了一下。
然后另一个 team lead 马上打电话二面,顺便考察一下是不是真是我自己写的程序。
这就是一天三面,两电一编程,然后就给 onsite了。面了组里的4个lead,HR 和
Founder。前两个基本都在问我的 research和 big ... 阅读全帖
x****3
发帖数: 62
21
刚拿到书, 还没看. 题是从http://www.crackingthecodinginterview.com考的. 感觉跟第4版差别不大.
Chapter 1 Arrays and Strings
1.1 Unique Characters in String
1.2 Reverse String in C
1.3 Check Permutation
1.4 Replace Spaces
1.5 String Compression
1.6 Rotate Image / Matrix
1.7 Set Row or Column to 0
1.8 Check Rotation Using isSubstring
Chapter 2 Linked Lists
2.1 Remove Duplicates
2.2 Find kth to Last Element
2.3 Delete Node from Middle
2.4 Partition List
2.5 Add Two Lists
2.6 Get Front of Loop in Circular List
2.7 Check ... 阅读全帖
x****3
发帖数: 62
22
刚拿到书, 还没看. 题是从http://www.crackingthecodinginterview.com考的. 感觉跟第4版差别不大.
Chapter 1 Arrays and Strings
1.1 Unique Characters in String
1.2 Reverse String in C
1.3 Check Permutation
1.4 Replace Spaces
1.5 String Compression
1.6 Rotate Image / Matrix
1.7 Set Row or Column to 0
1.8 Check Rotation Using isSubstring
Chapter 2 Linked Lists
2.1 Remove Duplicates
2.2 Find kth to Last Element
2.3 Delete Node from Middle
2.4 Partition List
2.5 Add Two Lists
2.6 Get Front of Loop in Circular List
2.7 Check ... 阅读全帖
y*******g
发帖数: 6599
23
git bisect 找出问题的code算不算?
g**********t
发帖数: 475
24
显卡用的是tesla M2070 (Fermi generation)。已经看过了cuda SDK里面的例子,但是
那个是bisection algorithm,貌似这个算法只能算eigenvalues,不能算eigenvectors?
c******n
发帖数: 100
25
来自主题: JobHunting版 - Facebook 今天的电面
今天的电面,payment 组的白女
1, valid palindrome, leetcode 原题,但是不能修改String, 所以就不能用s.
toLowerCase(), 当时就急了,因为不知道Character的 toLowerCase的method是什么,
她说写sudo-code就可以了。很快做出来了,然后写test cases
2, Binary Search的变型,Git Bisect, 从某一个版本开始,引入了一个bug, 然后让
我找出引入这个bug的第一个文件。binary search之
F家一般有几个电面啊,最近真的没什么时间。希望直接给onsite吧。
D*********G
发帖数: 193
26
来自主题: JobHunting版 - Facebook 今天的电面
你可以写个tolowcase method吧,一行?
Git Bisect 考的还挺多的, 主要看解决问题思路吧
还是那句话,想好test case,保证无bug
g********t
发帖数: 53
27
来自主题: JobHunting版 - Facebook 今天的电面
不能修改String,能否可以新开一个new String吧,在new String上操作?如果不行的
话, 就每次比较的时候, check 大小写。应该就可以
Git Bisect是什么题?愿闻其详
w****a
发帖数: 710
28
来自主题: JobHunting版 - Facebook 今天的电面
git bisect应该就是这题
http://lintcode.com/en/problem/first-bad-version/
z******f
发帖数: 277
29
来自主题: JobHunting版 - Facebook 今天的电面
Git Bisect不是个对半查的递归吗?
b*******w
发帖数: 56
30
来自主题: JobHunting版 - Google 电面
def sort_by_quadratic_poly(nums, a, b, _):
pivot = -(b/(a*2.0))
index_right = bisect.bisect_left(nums, pivot)
sorted_by_quadratic_poly = []
index_left = index_right - 1

while index_left >=0 and index_right < len(nums):
if abs(nums[index_left] - pivot) <= abs(nums[index_right] - pivot):
sorted_by_quadratic_poly.append(nums[index_left])
index_left -= 1
else:
sorted_by_quadratic_poly.append(nums[index_right])
in... 阅读全帖
h******v
发帖数: 36
31
来自主题: JobHunting版 - Google 电面
多用点space。
def sortys(arr, a, b, c):
from bisect import bisect_left as bl
ys = [a*x**x+b*x+c for x in arr]
x0 = -b/2/a
ri = bl(ys, a*x0**x0+b*x0+c)
li = ri -1
_ys = []
while li >= 0 and ri < len(arr):
if ys[li] <= ys[ri]:
_ys.append(ys[li]); li -= 1
else:
_ys.append(ys[ri]); ri += 1
_ys.extend(reversed(arr[:li+1]))
_ys.extend(arr[ri:]))
return _ys
a**6
发帖数: 4
32
来自主题: JobHunting版 - 这题有沒有P解?
不是找number of sequence , 是找ascending order of given sequence in all
sequences, 例如 4 9 7 7 有 12 个组合, 要找出4977在12个组合中的顺序排名(第3名
),期间也想过用bisection, 但是tree不是对称
a*******o
发帖数: 699
33
来自主题: PennySaver版 - 79个包子庆祝79(结束了)!
几何
bisect 平分
abscissa 横坐标
curve 曲线
topological 拓扑的
inclination 倾角;斜角
volume 体积
eccentricity 离心率

virology,
a*********n
发帖数: 602
34
Two World's Fairs
Main articles: 1939 New York World's Fair and 1964 New York World's Fair
The 1,255 acre (5 km2) park was created from the former dumping ground
characterized as "a valley of ashes" in F. Scott Fitzgerald's The Great
Gatsby. The site, known at the time as the Corona Ash Dumps, was cleared by
Parks Commissioner Robert Moses, in preparation for the 1939-1940 World's
Fair. Faced with having to dispose of the mountains of ashes, Moses
strategically incorporated a significant portion... 阅读全帖
M********4
发帖数: 1272
35
来自主题: Texas版 - 奥斯丁360这段路风景不错?
有何难哉?
1. FM 337, Hill Country, Medina to Camp Wood
2. FM 4, North, Jacksboro to Cleburne
3. State Highway 16, Bisects State, Antelope to Zapata
4. FM 1431, Hill Country, Round Rock to Lake Buchanan
5. Capital TX Highway Loop 360, RM2244 to RR2222
6. Lime Creek Road, Hill Country, Nr Cedar Park to Austin
7. FM 335, Hill Country, Camp Wood to SH 41
8. FM 3090, Southeast, Navasota to Carlos
9. FM 170, Big Bend, Study Butte to Presidio
10. FM 118, West, Kent to Study Bu... 阅读全帖
M********4
发帖数: 1272
36
来自主题: Texas版 - 奥斯丁360这段路风景不错?
有何难哉?
1. FM 337, Hill Country, Medina to Camp Wood
2. FM 4, North, Jacksboro to Cleburne
3. State Highway 16, Bisects State, Antelope to Zapata
4. FM 1431, Hill Country, Round Rock to Lake Buchanan
5. Capital TX Highway Loop 360, RM2244 to RR2222
6. Lime Creek Road, Hill Country, Nr Cedar Park to Austin
7. FM 335, Hill Country, Camp Wood to SH 41
8. FM 3090, Southeast, Navasota to Carlos
9. FM 170, Big Bend, Study Butte to Presidio
10. FM 118, West, Kent to Study Bu... 阅读全帖
c**********i
发帖数: 43
37
来自主题: Texas版 - 房产投资移民的项目请看看
我亲戚要投资移民, 项目在Dallas和Houston downtown, 我没去过texas,希望在当
地的大侠帮忙看看, 风险高不高, 值不值得投, 多谢了:
1. Uptown Dallas – Dallas, Texas Uptown Dallas, also known as “Lemmon/Oak
Grove” is a proposed Class-A+ multi-family development to be constructed on
a 1.07 acre site located in the heart of Dallas’ Uptown district. The
project site is situated along Lemmon Avenue and Oak Grove Avenue between US
75 and one block east of McKinney Avenue, the primary thoroughfare that
runs through Uptown.
2. Block 365 – Houston, Texa... 阅读全帖
p********y
发帖数: 5141
38
来自主题: Cycling版 - The Rules
Someone asked what the rules of bike jersey, and someone else point him to
this link. It is hilarious. Totally loved it! =D
------------------------------------------------------
THE RULES
http://www.velominati.com/the-rules/
by The Keepers / Jun 1 2009 / 6,291 posts
We are the Keepers of the Cog. In so being, we also maintain the sacred
text wherein lie the simple truths of cycling etiquette known as The Rules.
It is in our trust to maintain and endorse this list.
For those struggling to under... 阅读全帖
f*********n
发帖数: 11154
39
来自主题: GunsAndGears版 - 测试各位的瞄准功力(游戏)
有意思。
Your inaccuracy by category:
Parallelogram 3.6 7.0 1.0
Midpoint 1.0 3.6 2.8
Bisect angle 3.9 1.4 3.6
Triangle center 5.6 1.0 6.3
Circle center 4.1 4.1 1.0
Right angle 0.9 0.4 2.2
Convergence 2.8 2.0 5.1
Average error: 3.02 (lower is better)
Time taken: 135.0
p*********e
发帖数: 32207
40
来自主题: GunsAndGears版 - 测试各位的瞄准功力(游戏)
Your inaccuracy by category:

Parallelogram 6.3 3.2 5.0
Midpoint 2.2 1.0 1.4
Bisect angle 0.9 0.4 1.0
Triangle center 2.9 1.9 3.4
Circle center 2.2 3.2 2.8
Right angle 1.1 1.2 1.2
Convergence 1.0 2.2 2.0
Average error: 2.21 (lower is better)
Time taken: 427.7
记录里面有一个1.45, 76秒的,简直是畜生啊.
p*********e
发帖数: 32207
41
来自主题: GunsAndGears版 - 测试各位的瞄准功力(游戏)
快速瞄了一次
Parallelogram 4.5 2.0 2.2
Midpoint 1.0 0.0 1.0
Bisect angle 2.8 2.5 1.1
Triangle center 4.4 8.6 2.3
Circle center 3.2 2.8 1.4
Right angle 2.6 5.1 1.8
Convergence 2.2 3.2 3.2
Average error: 2.76 (lower is better)
Time taken: 97.4
难以想象76秒内搞定1.45的驴...
N*****7
发帖数: 1899
42
来自主题: Tennis版 - 签表出来了
Probably not conincidence, but ranking predetermined grouping.
Rankings of #1-#6 have not changed names this month,
neither does grouping.
#1 #4 same group
#2 #3 same group
so on
Actually even if #2/#3, #6/#7 swap name,
it still ends the same bisection.
b*********s
发帖数: 6757
43
来自主题: Tennis版 - 再奔一个LONG GAME
终于静下心看了lz的和其他大侠们的意见。我非常认同 WatermanCT 的意见, 当我看时
,这也是我想说的意见之一(he beat me to it). it's always easier to move
forward to hit a ball than falling back.
stroke 而言, 不容易说, 你有些球的stroke 有问题是因为你站太前, 对手球深,
你一时来不及而造成的, 特别是反手。 有几个正手打得很不错。 准备最好能够再早
些。 要看stroke的话, 这个角度不是最佳。。。 真要analysis 动作的话, 你换个
camera 角度, 让人喂球。 不过lz的stroke 不是什么特别大的问题。
cross court 问题, 用 semi-western, western grip, 打cross court 都有些
tricky. 看了你几个cc, especially on the run, 一个普遍的问题(including
myself),是stance. 你基本是右脚在左脚后方relative to the baseline.... 阅读全帖
b*********s
发帖数: 6757
44
it should be semi open stance. Keep in mind, because of you are running
around the center of the bisect angle is no longer directly across the net (
"center" of the court is now moved further to opponent's ad court, and
because of this it may look a bit like neutral stance.
a*******e
发帖数: 12169
45
Your inaccuracy by category:

Parallelogram 2.2 5.0 5.0
Midpoint 2.0 2.0 7.6
Bisect angle 0.4 5.1 1.0
Triangle center 9.6 2.1 2.4
Circle center 2.0 1.0 2.8
Right angle 1.2 1.1 1.7
Convergence 8.1 5.0 8.5
Average error: 3.61 (lower is better)
Time taken: 77.6
b***n
发帖数: 590
46
Your inaccuracy by category:
Parallelogram 6.1 6.3 3.2
Midpoint 0.0 3.0 3.2
Bisect angle 1.2 2.3 1.6
Triangle center 2.9 2.3 4.1
Circle center 2.2 5.0 1.0
Right angle 0.8 1.7 3.4
Convergence 3.6 3.6 1.4
Average error: 2.80 (lower is better)
Time taken: 303.3
x*****g
发帖数: 2118
47
挺好玩的
Your inaccuracy by category:

Parallelogram 2.0 1.4 5.0
Midpoint 3.6 1.0 1.0
Bisect angle 1.9 4.0 0.3
Triangle center 5.4 2.9 1.2
Circle center 5.0 5.1 3.0
Right angle 2.2 0.4 0.5
Convergence 2.2 4.0 4.2
Average error: 2.68 (lower is better)
Time taken: 163.0
f*******l
发帖数: 8811
48
【 以下文字转载自 Military 讨论区 】
发信人: Yossarian (野马尘埃), 信区: Military
标 题: 我党最神奇的间谍,男扮女妆,深入法国
发信站: BBS 未名空间站 (Thu Nov 3 10:47:38 2011, 美东)
这是我几年来看到的最娱乐最搞笑的故事。
不知大家看过M. Butterfly这部电影没有。
汉语翻译蝴蝶夫人,蝴蝶君。
http://www.imdb.com/title/tt0107468/
这部电影基本上是写实的,讲述了时佩璞的传奇故事。
一下引百度百科
’时佩璞,男,是北京青年京剧团的编剧、旦角演员兼团部秘书。时佩璞毕业于云南大
学,主攻法语兼西班牙语,学生时代就喜欢京剧,曾与关肃霜合作演出。他后来拜小生
姜妙香为师,曾经在北大礼堂演出《奇双会》。
1964年认识当时驻北京的法国外交官Boursicot,对方以为他是女儿身, 猛烈追求他,
两人其后更发生性关系。在北京工作的布尔西科,一次在使馆举办的酒会期间遇上男扮
女装的花旦演员时佩璞,两人陷入爱河,但布尔西科一直都不知道时佩璞是男儿身,两
人密切交往后,时佩璞更告诉布... 阅读全帖
u****q
发帖数: 24345
49
来自主题: Joke版 - 公司里正在写代码的妹子
git init
git bisect
git merge
git pull
git push
...
git commit
git reset
d****o
发帖数: 32610
50
来自主题: Joke版 - Re: 法国科学家 (转载)
【 以下文字转载自 Military 讨论区 】
发信人: Xujiahui (南模是所好学校), 信区: Military
标 题: Re: 法国科学家
发信站: BBS 未名空间站 (Sun Apr 24 00:25:31 2016, 美东)
谢谢,我没干过数值计算的活,这次是赶鸭子上架。我大学是在法国读的,所以有些中
文术语不清楚,牛顿迭代是不是就是newton raphson?收敛应该是converge吧?
我的问题有点复杂,前面没有解释清楚。关键是逼近值一定要准,因为这只是一个单元
里的逼近值,然后输出到下一个单元作为输入。一共有几千个单元,所以如果不准的话
最后的误差会被放大。我最开始用的是brute forcing,一步一步向交点逼近,计算停止
在误差小于0.1%(举例),但是如果我的step取的不够小的话,可能最靠近交点的逼近
值都不能在0.1%以内,无法converge。但是step取的小的话,计算量又很大。
bisection我也用过,也不准尤其是在多个单元之后,误差会越来越大。
我还试过shooting method,其实差不多就是newton raphson了... 阅读全帖
1 2 下页 末页 (共2页)