n*w 发帖数: 3393 | 1 java的generic不是宽松。是公认的poor implementation。见过一人对list,list
重复写几次。理由是避免java generic的box/unbox的performance问题。
可重用代码很多方式,有部分是要type system的支持。 |
|
g*****g 发帖数: 34805 | 2 现在又不要谈实时trading了?很好,那你就来说说到底什么系统boxing/unboxing居然
能成为系统的瓶颈。我老给的link,是一堆常见的小算法程序shootout,咋就没有一个
把这个瓶颈体现出来呢?
我老要是谈Strong type,会去举web service这种例子,而不是boxing和性能问题。你
成天驴唇不对马嘴被打脸是必然的。学了个helloworld,看到某个blog自吹Android上C#
比java快7倍,就以为拿到令箭了。 |
|
p**o 发帖数: 3409 | 3 到处声明类型的确比较臃肿;但全部duck typing很难做local reasoning,我大脑常常
栈溢出。这跟写代码的人的水平关系很大。高质量的Python代码读起来比高质量的Java
代码愉快十倍;但Python烂代码读起来比Java烂代码痛苦一百倍,常常只能推倒重写。
我觉得在函数签名上有选择地做类型标注或者声明、函数内部不做声明,是比较好的折
中。
Python3的function annotations特性很受欢迎,虽然只是标注而没有影响语义。mypy
是个有意思的python编译器和虚拟机,利用这个特性做static typing,也就是说在语
法跟python(基本)相同,但语义上是python的超集。
http://mypy-lang.org/
( |
|
p**o 发帖数: 3409 | 4 因为现代虚拟机的发展,compile time和run time的区分已经越来越模糊了,所以以此
为区分的static typing和dynamic typing也在融合。在实际使用中,一个需要在
runtime才定类型、但可以经由JIT充分加速的语言,与一个虽然在compile time可以由
机器推导定类型、但平均水平的人脑很难在局部迅速做出类型判断的语言,在运行性能
、代码可读性、重构的灵活性这三个实质性指标上没有什么不同。
dynamically |
|
b***e 发帖数: 1419 | 5 var TC = require('web-service-api').TypeCheck;
var tc = new TC({
typeDefs: typeDefs
});
tc.typeCheck(value, type);
This type check part of this package is light-weight, standing alone, and is
not tied to the web service part. |
|
p*****2 发帖数: 21240 | 6 type class 用处大吗?
haskell可以理解 scala有trait又有implicit class type class比haskell麻烦很多
感觉基本没啥用呀 难道就为了学haskell而学? |
|
p*****2 发帖数: 21240 | 7 Type classes are useful to solve several fundamental challenges in software
engineering and programming languages. In particular type classes support
retroactive extension: the ability to extend existing software modules with
new functionality without needing to touch or re-compile the original source
. [1]
这个implicit class就可以实现呀 而且更直观 |
|
|
gw 发帖数: 2175 | 9 这样定义了数据类型
#ifdef USE_DOUBLES
typedef double dataType;
#else
typedef float dataType;
#endif
然后在程序中如果出现
std::complex x,y;
x=complex(2.0,0);
y = 2.0*x
就会有这样的问题:
Type no match for 'operator/' (operand types are 'std::complex' and '
double')
如何破? |
|
m****o 发帖数: 182 | 10 tensorflow的java API实在太原始,而且还不能训练,无奈之下捡起了Python。写了段
时间发现duck typing is surprisingly good and productive,再加上namedtuple作
为immutable case class/record,还有type annotation,对于Python的整个印象大为
改观。 |
|
e*******e 发帖数: 248 | 11 刚写的会议文章,上IEEE的pdf testing site检查capatibility。
结果说用了type 3 font,不能submitt。我花了好长时间才找到罪魁祸首:pstrick的
psdot命令。将所有psdot换成pscircle后,就一切ok了。不知道还有那些类似的命令使
用了type 3 font? |
|
c****r 发帖数: 185 | 12 我用MiKTeX 2.5,latex->dvi->ps->pdf。
结果产生T3font_5 的Type 3 font。
如何设置成只产生type 1 fonts?
多谢。 |
|
g****z 发帖数: 90 | 13 I use PC (XP) + x32 to connect a remote linux system. I run xterm on my XP.
Suppose I have a file "abcdedfghijklmn.txt".
I want to type "cp abcdedfghijklmn.txt".
I have to tpye every character and the autocomplete does not work. i.e.
If I type "cp abc[tab]", it does not work.
How to get this function?
Thanks a lot. |
|
w********e 发帖数: 275 | 14 请问哪位了解高精度HLA Typing 的市场应用。似乎有一些血库需要,但不知为什么?
器官移植是不是不需要太高的HLA Typing精度?
Thanks! |
|
f****l 发帖数: 14 | 15 【 以下文字转载自 Medicine 讨论区 】
发信人: forall (我的黑夜~记忆的马甲), 信区: Medicine
标 题: 请问在哪做hla typing examination?
发信站: BBS 未名空间站 (Thu Jan 7 20:47:35 2010, 美东)
需要做一个 HLA Typing Examination. 在一般的医院能做吗? 还是需要去专门的医院
? 一般的诊所给做吗? |
|
A****F 发帖数: 699 | 16 As many of us know our "Water Cube" was using this type of structure and
designer said this type of structure is the most efficient way to divide
space. This is in fact the same way that bubbles form in foam.
Whereas soap bubbles can be divided into two shapes - three quarters of the
cells have 14 sides and the remainder has twelve - the engineering solution
to make the Water Cube a reality requires over a hundred different ones.
However, in spite of their seeming randomness, bubbles always touc |
|
A****F 发帖数: 699 | 17 应该是2小时。除非是type 2,3,4,5才允许1 hour fire rated substituted by
sprinkler. But for type 1, no substitution.
★ 发自iPhone App: ChineseWeb - 中文网站浏览器 |
|
O*A 发帖数: 78 | 18 【 以下文字转载自 Physics 讨论区 】
发信人: OSA (Opt Soc Am), 信区: Physics
标 题: 请问ZnO能否用作p-type Si的透明电极
发信站: BBS 未名空间站 (Sun Jun 20 15:48:20 2010, 美东)
在p-type Si上面沉积ZnO做透明电极是否可行,会不会形成pn结?
Si表面重掺杂 |
|
g********m 发帖数: 956 | 19 【 以下文字转载自 Science 讨论区,原文如下 】
发信人: glassdream (dashao), 信区: Science
标 题: Type 3 Font Error when WORD --> PDF
发信站: Unknown Space - 未名空间 (Tue Mar 1 11:35:28 2005) WWW-POST
该死的会议论文提交系统,要求每个PDF文件必须通过它的检查方可上载。
我是在MS WORD里编辑的公式,
每次通过以下操作生成PDF文件,
print to file (.prn) --> Gsview (.ps) --> Adobe Distiller (.pdf)
该系统就检测出一堆Type 3 Fonts errors.
我得一个一个公式重新改写。
但现在,它死活不认WORD里的积分号字符(以及MS WORD EQUATION EDITOR下的积分符号)
,
有没有大侠给我指点一下,如何让积分符号在被编译的PDF文件中通过检查。
或者用什么其他符号可以替代积分?
多谢。 |
|
i*******y 发帖数: 225 | 20 【 以下文字转载自 EE 讨论区 】
发信人: iluvTOfly (sky is not the limit), 信区: EE
标 题: 那有卖release type solenoids?
发信站: BBS 未名空间站 (Mon Nov 17 11:24:11 2008)
有一个project, 需要hold一个1lbs左右的重物(可以用绳子拴住重物),
通电后自动release重物。我想release type solenoids应该可以实现。
但找不到哪里有卖的,难道要自己用solenoids加工?
大家支支招,谢谢 |
|
l********y 发帖数: 2295 | 21 can u type * ? that's we usually typed for multiply. might be wrong. |
|
r****r 发帖数: 29 | 22 I was told that the computer function of PN writing is very limited. For
example, after finishing typing a sentence, if I want to change a word
within this sentence, I have to erase the whole sentence, and then type from
beginning of the sentence.
Is this true? Please advise. |
|
m*****3 发帖数: 56 | 23 When I practice typing PN in USMLE website, the response always gets delayed
. it only happens in my PC laptop, not for desktop or Mac. Does any one
know how to fix the problem? I can only practice PN typing at home with
laptop.
Thanks and best wishes for Chinese New year ! |
|
n*********d 发帖数: 87 | 24 It's a bit weird for me to type in English here. However, I don't have a
software to type in Chinese. Would anyone recommend one for me to download?
Many thanks. |
|
v*******e 发帖数: 11604 | 25 不知道版上有没有人做量子光学的。我不是做这个的,只是业余爱好。现在有个问题思
考不出来,向知道的人请教下,谢谢了。
那个BBO晶体产生纠缠光子对,在量子光学上用得很多,早期也被人用来做Bell test。
比如这个文章,PRL的,new high-intensity source of polarization-entangled
photon pairs,网上有免费pdf下载的。用氩粒子UV激光器照射BBO晶体的时候,由于非
线性作用,这个晶体会产生一些低一点频率的光子对。其中有一类是两个光子相等频率
的,其中一个光子的偏振和入射激光一致(ordinary),另外一个光子(extraordinary)
的偏振是和入射激光垂直,并且这两个光子的偏振是相互纠缠的。O光子的出射在一个
cone形状上,E光子的出射也在一个cone形状上,如果晶体光轴切得合适,这两个cone
形状之间有两条相交的直线。一对纠缠的光子它们的出射角分别在两个cone上,但是他
们是对称的。
看到的文章都说,如果选择沿着两条相交的直线射出来的一对光子,他们不仅是相互纠
缠的,而且这一对光子处于叠加态,记作|H1,V... 阅读全帖 |
|
w***j 发帖数: 11 | 26 弱弱的问一个,还有type 1 analysis 吗?
他们都是做什么用的?
先谢谢了 |
|
e****t 发帖数: 766 | 27 in proc phreg, the results from "type 3 test " will change if i change the
coding method. (1 vs 0 or 1 vs 2) , i put the variable in class statement
already. ideally, it will be the same , as I consider as categorical data.
. the coding wont have impact...
(the analysis of MLE part will change as the coding change will leave to
main effect definition changes.. dont know why type 3 changes too.. )
anybody knows the reasons ?
Many thanks in advance ? |
|
w*****n 发帖数: 375 | 28 Test的critical value 是你自己来定, 所以你可以调整critical value 来让 type I
error rate 正好在 nominal level.
比较两个test的时候, 调整两个的critical values 来让type I error rate 都在
nominal level的时候才是公平比较。 |
|
g******2 发帖数: 234 | 29 depends on your standard error of your empirical type 1 error. If your
empirical type 1 error is significantly higher than 0.05, you probably need
to make your threshold more strict. |
|
|
W**********g 发帖数: 3340 | 31 A dating expert encourages women to ditch their checklists and be open to me
eting men outside their comfort zone.
- Delaina Dixon, BettyConfidential.com
"He's just not my type!" If that's the exclamation that comes out of your mo
uth every time you meet a guy you like but he doesn't fit your “dream man”
criteria, and you're still single, Andrea Syrtash would like to have a word
with you.
The dating expert has written a new book, He's Just Not Your Type (And That'
s A Good Thing), and in it, she |
|
m*********8 发帖数: 1018 | 32 从05年开始口碑还是不错的.具体的当然是你自己去X-Type forum潜水最方便.
这里就隆多开X-Type吧 |
|
F*********e 发帖数: 2696 | 33 这代typr r是唯一一代美版type r,和日版type r几乎一样的 |
|
c**u 发帖数: 2230 | 34 different animals. TL是让你开着享受的同时来点sporty的flavor. type R是为
racing 和赛道而生的, 当然是减重了, 好象有一款连AC, 后窗的wiper都给卸了, 很纯
粹的.
至于civic si, 是后来有compromise的. 没有type R那种纯粹的bloodline. |
|
k***x 发帖数: 6799 | 35 你这个说得不错,其实我现在的02 tl-s才买了半年,开得还是好好的,唯一不爽的就
是,一个大妈买菜车mpg都只能到17,我实在无语了
我看了一下integra type r,居然都能到25/30,而且是coupe(能坐4人吧?),实用
性也还可以
s2000的话性能肯定没得说,但是油耗可能比tl-s还差,而且不实用,我每周买个菜估
计都装不下……
civic si以前我以为不咋样(菜鸟),听诸位一说好像还很牛叉,不过主要是开dfbb的
太多了,开个玩笑,我在学校停车场扔块石头都能砸中好几辆civic
所以一折衷,我觉得integra type r还不错,不过大家说得有道理,这个价格太夸张,
而且还得把tl-s给卖掉,小有点麻烦;不怕大家见笑,我是在这边学车的,拿到驾照还
不到2年,连手动车都不会开,看来还是算了。
反正最近正好要给tl-s做12万维护,看看换了火花塞之类的会不会好一点。Acura论坛
上有人说给车清洗了egr ports,换了火花塞,再seafoaming之类居然mpg到了33,我还
是小有点激动的,大家说说这个靠谱否?
http://tl.acurazine.com... 阅读全帖 |
|
e****x 发帖数: 148 | 36 之前本田官方有confirm过会把type-R上那台vtec 2.0引进美国市场
这次又官方宣布引入type-R,如果换个发动机doesn't make sense... |
|
s********u 发帖数: 1109 | 37 同样价位z06比F-type R不知道快到哪里去了,马力更大,还轻得多。当然F-type就是
GT car,z06已经比较hardcore了 |
|
|
|
|
|
|
C***r 发帖数: 759 | 43 The Ideal Type
In his effort to escape from the individualizing and
particularizing approach of German Geisteswissenschaft and
historicism, Weber developed a key conceptual
tool, the notion of the ideal type. It will be recalled that
Weber argued that no scientific system is ever capable of
reproducing all concrete reality, nor can any
conceptual apparatus ever do full justice to the infinite
diversity of particular phenomena. All science involves
selection as well as abstraction. Yet the social |
|
W*****2 发帖数: 1043 | 44 Obama's speech of the community-organizer type
After assisting Hilliary Clinton in the White House for three years, Obama
is still at the level of a community organizer. Nothing concrete except
empty promises, read my lips and hype
When will he blame himself of being unable or unwilling to get jobs done????
??
Just wondering when will he ever improve????? |
|
l******r 发帖数: 18699 | 45 how to switch the keyborad to Chinish typing? |
|
l******r 发帖数: 18699 | 46 nobody knows how to type Chinish in macbook? |
|
|
|
d*******d 发帖数: 2050 | 49 这辆车买了半个多月了,1400mile了,可以有个详细全面的评估了.
2006 acura, rsx, type-s, 6-speed manual.
price, $22000 + tax,doc.......
每天单程30mile,其中2mile local, 20mile平地高速,8mile山地弯道.
8mile山地,4mile上山,4mile下山,落差1500feet.每天一个往返,大约一
共60mile.
买这辆车考虑的是:
1. 需要一量commute car
2. 要有点fun.
3. 不要太贵.
4. 烧油不要太多.
5. 必须是手动.
1. powertrain的感觉
当然,这一辆比上个del sol强多了.不过,torque还是让人不太满意.
torque的输出曲线基本上是/----\高原型.从大约3500转到8000转的torque
基本上稳定的.好处是,在高速上,80mile时,还能在六档上获得有力的加速.
可是,在低端时,torque不足,加速不够有力.0速启动时,就算是在2档上
踩到5000,6000,7000转,也只有soso的加速度.(其实4000-50 |
|
y***a 发帖数: 3 | 50 需要order DPS Type 3A Drive Record.登陆网站按照要求填写信息。除了驾照号,出
生年月,
adult No.,还需要填SSN末四位。但是我是F2,没有ssn。这种情况下该怎么办?请牛人
指教。 |
|