f******e
发帖数: 2888 | 1 那是不一样的。
不知如何描述，我女儿也是不到2岁就可以从婴儿床爬出来，但是到现在快3岁了，还没
有从safety gate里爬出来过。---一个圈或者是房间的门。
我邻居4岁儿子也出不来（也是活动能力很强的），但是5岁多的那个儿子可以越过
saftegate
而且，saftety gate还有更高点的。
... |
|
l*x
发帖数: 14021 | 2 可能因为二线的轮换少。DL和LB的轮换比较频繁。不同的人打不同package。
主力CB和Saftety换的少。绿湾的保安Morgan Burnett，去年参加了一整赛季所有的
1000＋ snap。今年一直受伤，严重影响绿湾二线实力。对三番被保定完爆。所以看
tackle总数的话，很可能二线的球员偏多。
好的二线球员不好找，基本就是主力那几个。如果受伤的，可能二线实力影响很大。所
以导致成绩差。二线主力稳定的话，防守就好打不少。 |
|
s*****h
发帖数: 44903 | 3 Alexander Hamilton
Several monumental Dallas mistakes:
1) Bad clock management on the final drive. Jason Garret wasting a down by
spiking the ball on the final drive. You do not want leave any time on the
clock for Aaron Rodgers. Lack of a sense of urgency to score a touchdown
versus a field goal.
2) Overreliance on Witten and not enough Dez Bryant on the final drive.
three straight passes to a TE on final drive...think about what T.O. said
years ago about the Cowboys focusing too much on Witten... 阅读全帖 |
|
l******u
发帖数: 3169 | 4 谢谢大家的信息！如果有club或者meet-up的确也是比较好的选择，主要是刚发现REI居
然有这种trip，我的理解是它会提供比如炉子帐篷救生用品之类的设备，至少可以
rental，可以在以后采购时少走一些，同时也是一种体验。我最终的目标是可以带着家
人去一些这样的trip，希望能多学一些东西让她们可以更加好的享受backpack camping
的乐趣。
car camping和day hike我还是有些经验，只是真的backpacking还是觉得千头万绪不知
从何开始 ：-）
另外发现REI还提供两天的wilderness saftety class，不知道有没有人参加过？我们
这里是小地方，刚刚开了个REI，我还在新鲜中。。。hehe |
|
t******a
发帖数: 1200 | 5
1. 你知道什么是 double buffering 么? 你看过 Java Stream 的实现么? 你
Benchmark 过Java IO Library vs C library 么 ? 不要整天拿个你不懂的名词
跑火车。 Java 做 serialization时因为 type saftety 原因不支持类似
readFoo( &myfoo, sizeof(Foo)) 的方法，但单纯的 read, write 是不慢
的， 尤其是 disk/network io.
Premature optimization is the root of all evil. - Don Knuth
Rule 1: Don't optimize your code. Rule 2 : Still don't optimize your code
用 C 实现相同的功能速度不可能是 Java 的十倍。按照我们的经验，Java 比 C
一般情况下慢 20% - 100% 左右。如果有特别 computational bootleneck,
用 JNI 写一个 Native implementat |
|
|
r*****o
发帖数: 140 | 7 我的理解是可能一样大，但是有时候safety可能会更大一点。
关键是看protocol。
因为有可能受试者在随机化后虽然给过药，但是没有任何疗效评价,protocol认为需要
剔出FAS（即ITT）；或者受试者严重违背入组标准，按照protocol的要求不进入FAS分
析；或者发现某个重要的计算主要疗效指标的基线数据缺失，无法算疗效。
但只要以上情况的受试者被给了药，仍然可以进入Safety分析。 |
|
m****r
发帖数: 202 | 8 A safety population
is a group of patients who have taken the drug at least once.
An ITT population
was defined as those individuals who had applied their assigned treatment at
least once and had returned for assessment of the primary end point (such
as healing of the fissure). Patients who failed to present for assessment
and who would not reattend for assessment to determine an outcome were
deemed to have withdrawn their consent to participate and were excluded from
the analysis.
In our work, ... 阅读全帖 |
|
d***2
发帖数: 341 | 9 This is usually called MITT (Modified ITT), the true ITT is the population
of subjects who are randomized or enrolled (if not randomized trials) to
receive study treatments.
To answer LZ's question on population size:
ITT >= Safety >= MITT >= Per Protocol (this depends on the definitions of
major protocol violations)
Of course, there could be odd cases that change this order, for example, you could get a non-randomized subject who receives study medication. Don't ask me how, it happens.....
at
... 阅读全帖 |
|
t********i
发帖数: 7856 | 10 MDX is good.
BTW, if saftety is your top priority, consider Touareg, but it does not save
gas. |
|
