l*******g
发帖数: 82 | 1 那个水滴下繁殖的可不可以用logrithm算以m为底log1000？
log 1000 / log m (都是10为底)
数据库查找有没有，应该加一个cache，只有其他修改方法调用cache才会清空无效。 |
|
l*******g
发帖数: 82 | 2 上面说错了，水下那个等于是minutes ~= log (1000/m) base 2
不知道是不是？简单的来说就是2^minites * m = 1000
那个水滴下繁殖的可不可以用logrithm算以m为底log1000？log 1000 / log m (都是10
为底)数据库查找有没有，应该加一个cache，只有其他修改方法调........ |
|
f*******s
发帖数: 182 | 3 O(1) = constant time/space, oh-one
O(logn) = logrithmic time/space, log-n
O(n) = linear time/sapce, oh-n
O(nlogn) = n-log-n
O(n2) = quadratic time/space, n-squared
O(n3) = n-cubed
O(2n) = exponential, two-to-the-n
O(n!) = n factorial |
|
o**y
发帖数: 3065 | 4
1992
庄。
Actually Chinese leaders have been so smart, they knew that no country will
be the single dominance of the world in the next century, the world politics
will be in a multipole framwork. US is still the dominance of North America
, Brazil should be the dominance of South America, Germany and France are
both dominance of Western Europe, Turkey will show great potential in middle
east, China will be dominace of Far East. Japan will compete with China,
but Japan has serous demographic probl... 阅读全帖 |
|
n******r
发帖数: 989 | 5 这个我能信，我觉得如果QE3真的来了，这个预测有潜力可挖。但这不是太悲剧了吗？
坊间有这样一个说法，经济萧条的时候，富人会希望从黄金这上面来达到保值的有效性
。所以黄金飘高是对美国经济实体的反讽之镜。我认为美国从来都不是一个能引导世界
消费格局的国家，恰恰相反，法国才是！但现在是亚洲市场占据了主旋律。美国的创新
格局可能暂居首位，但推动市场活力和激活就业状况的受益者依然是亚洲那边的。所以
美国这地方，还真看不出未来五年有什么源动力能扩大就业规模。要知道人家都去亚洲
干事业了，美国的酒店就没什么生意人过来住了，没看那些五星级hotel coupon满天飞
吗。
if you plot the gold prices vs time, in a logrithm format, you can see the
gold price in this decade is increasing as a straight line, if you
extrapolate it into 2016, the price will be $5000/oz. If you take the CPI
... 阅读全帖 |
|
|
H****h
发帖数: 1037 | 7 Proof No.1
1 true for n=2;
2 true for n=2^k;
3 true for n if n< 2^k
apply 2 for b_j, 1\le j\le 2^k, where
b_j=a_j, 1\le j\le n; b_j=average of a, n+1\le j\le 2^k.
Proof No. 2
Use the fact that logrithm function is concave. |
|
o**a
发帖数: 76 | 8 the roots of cos(sqrt(z)) is (n*pi+pi/2)^2
and \sum 1/(n*pi+pi/2)^2 converges
so I set
cos(sqrt(z)) = g(z) * \prod (1-z/(n*pi+pi/2)^2)
and take logrithmic derivative on both side to solve for g(z)
but I finally got g(z) = 1/cos(sqrt(z))
What's the problem here? What's the genus of cos(sqrt(z)) indeed? |
|
B****n
发帖数: 11290 | 9 You can take logrithm to this product and use linear (or quadratic)
approaximation to check whether it converges or not. |
|
B****n
发帖数: 11290 | 10 y-axis 是不是應該take logrithm? |
|
p**********m
发帖数: 143 | 11 嗯当然。已经take了。
y-axis 是不是應該take logrithm? |
|
B*********h
发帖数: 800 | 12 ☆─────────────────────────────────────☆
realchris (chris) 于 (Thu Apr 19 13:40:23 2007) 提到:
证明
比较大小：3的pi次幂与pi的3次幂
☆─────────────────────────────────────☆
wusuowei (student) 于 (Thu Apr 19 13:43:26 2007) 提到:
take logrithm of each with base 3 ...
☆─────────────────────────────────────☆
realchris (chris) 于 (Thu Apr 19 13:55:26 2007) 提到:
still no result
☆─────────────────────────────────────☆
SYSUP (SYSUP) 于 (Thu Apr 19 14:00:25 2007) 提到:
Then?
I think we can consider functions f1 |
|
p******d
发帖数: 374 | 13 Sorry that I can not input Chinese now.
I think you have the following fomular: Y=Exp(-ct)
while c is the time constant
Then you can know that Ln(y)=-ct
It is simple now since the only thing you should do is get the logrithm
then linear regression.
You can know whether your data obeys the expontential formula
Gook Luck |
|
