w********u
发帖数: 328 | 1 i have one series of observation data and one series of expected data, i
want to know if they're significant to each other, so i use chi-square
method, x=sum(Oi-Ei)^2/Ei, it's pretty straightforward.
But there are different Ti (time interval) between every data and i want to
take them into account as well, i mean, i hope there is a time-interval
variable in the formula. so i think it has sth. to do with the time series,
which method could be used? any hint is OK, thank you!!! | D*********2
发帖数: 535 | 2
>> significant to each other? would you please be more specific what's ur
test hypothesis?
so i use chi-square
>> how to calculate Ei in ur case?
to
,
>> what's ur data looks like?
Oi Ei Ti_upperbound Ti_lowerbound
【在 w********u 的大作中提到】
: i have one series of observation data and one series of expected data, i
: want to know if they're significant to each other, so i use chi-square
: method, x=sum(Oi-Ei)^2/Ei, it's pretty straightforward.
: But there are different Ti (time interval) between every data and i want to
: take them into account as well, i mean, i hope there is a time-interval
: variable in the formula. so i think it has sth. to do with the time series,
: which method could be used? any hint is OK, thank you!!!
| s*****9
发帖数: 108 | 3 我所知道的大部分time series都是针对等距离的time interval的，对你这个问题帮助
不大吧。如果expected data是通过某个model/mechanism得到的，就可以在相同的
timeline上比较两组数据了。 |
|
