b*e
2 楼
顶。
l*r
7 楼
one line code in R
l*r
12 楼
you first need to smooth your data by local fitting (loess()), and use a
function from pastecs library to find the turning points, then compare the
position of turning points.
Well, not one line code, but it is pretty straight-forward. The tricky
thing is to tune the parameter.
【在 b*******e 的大作中提到】
: 大牛给个link! Orz
function from pastecs library to find the turning points, then compare the
position of turning points.
Well, not one line code, but it is pretty straight-forward. The tricky
thing is to tune the parameter.
【在 b*******e 的大作中提到】
: 大牛给个link! Orz
b*e
13 楼
Orz:)
【在 l*******r 的大作中提到】
: you first need to smooth your data by local fitting (loess()), and use a
: function from pastecs library to find the turning points, then compare the
: position of turning points.
: Well, not one line code, but it is pretty straight-forward. The tricky
: thing is to tune the parameter.
【在 l*******r 的大作中提到】
: you first need to smooth your data by local fitting (loess()), and use a
: function from pastecs library to find the turning points, then compare the
: position of turning points.
: Well, not one line code, but it is pretty straight-forward. The tricky
: thing is to tune the parameter.
b*e
15 楼
果然经验老道。拜谢!
【在 l*******r 的大作中提到】
: you first need to smooth your data by local fitting (loess()), and use a
: function from pastecs library to find the turning points, then compare the
: position of turning points.
: Well, not one line code, but it is pretty straight-forward. The tricky
: thing is to tune the parameter.
【在 l*******r 的大作中提到】
: you first need to smooth your data by local fitting (loess()), and use a
: function from pastecs library to find the turning points, then compare the
: position of turning points.
: Well, not one line code, but it is pretty straight-forward. The tricky
: thing is to tune the parameter.
b*e
18 楼
呵呵,那我给一个好了
Double tops ~DTOP! and bottoms
~DBOT! are characterized by an initial local extremum E1 and subsequent
local extrema Ea and Eb such that
Ea [ sup$Ptk
* : tk
* . t1
* , k 5 2, . . . ,n%
Eb [ inf $Ptk
* : tk
* . t1
* , k 5 2, . . . ,n%
and
DTOP [ 5E1 is a maximum
E1 and Ea are within 1.5 percent of their average
ta
* 2 t1
* . 22
DBOT [ 5E1 is a minimum
E1 and Eb are within 1.5 percent of their average
ta
* 2 t1
* . 22
B. The Identification Algorithm
Our algorithm begins with a sample of
【在 b*******e 的大作中提到】
: 你知我知天知地知,电脑不知怎么算。-_-
Double tops ~DTOP! and bottoms
~DBOT! are characterized by an initial local extremum E1 and subsequent
local extrema Ea and Eb such that
Ea [ sup$Ptk
* : tk
* . t1
* , k 5 2, . . . ,n%
Eb [ inf $Ptk
* : tk
* . t1
* , k 5 2, . . . ,n%
and
DTOP [ 5E1 is a maximum
E1 and Ea are within 1.5 percent of their average
ta
* 2 t1
* . 22
DBOT [ 5E1 is a minimum
E1 and Eb are within 1.5 percent of their average
ta
* 2 t1
* . 22
B. The Identification Algorithm
Our algorithm begins with a sample of
【在 b*******e 的大作中提到】
: 你知我知天知地知,电脑不知怎么算。-_-
c*y
24 楼
这就叫计算机算命!
相关阅读
像知道为什么DT容易输钱?Re: 高晓松:人类精华犹太人 (转载)PCLN 不认为会在这里盘很长时间下周有什么好买的?看看想想:啥是 CAFR ??? (转载)big bear lake,谁去过?受apple官司的启发,我们家LG要file专利UBNT北大版的版主疯了 (转载)BAC要是把Countrywide spinoff 那就热闹了樊纲炮轰郎咸平:4万亿投向无误 他不知基本事实 (转载)谁能给我几个长期最不被看好的公司的pick讨论:04~05年股票市场萎靡不正是什么原因?为什么不开对赌博彩,版务全都放假了吗听说过走尸,但实际上很轻易被制服what will happen to TUDO's shareholder?zz : 北大又出事了[实名举报]北京大学学生会主席戴威 (转载)想投资美股,在chase开户交易成本是不是很高?美国经济越来越糟糕下周一盖铺当,OK?