600 people's time series
Your questions are as basic that it smells like home/classwork. Try to
formulate precise questions and not questions like "how do I solve this
general problem" and you are much mor likely to get answers.
To get you started:
For questions on how to plot values from dataframes please have a look
either at "an introduction to R (pdf)", some basic textbooks or just
google: "r-project plotting tutorial" (the second link should get you
started)
For the regression I would loop through all ids and calculate the
regression for a subset of the dataframe like:
for (i:1:...)
{
invest = DF$invest[DF$id==i]
payout=....
invest.norm=invest/mean(...)
l.model=lm(...)
}
You will have to find out the details yourself. We will happily help if
you post some precise (!) questions in case you get stuck!
Jannis
On 04/16/2011 07:15 PM, ???Su Jiangdong wrote:
Hi there,
I have a data frame DF of over 600 people's short term trade data in time
order. Below is the super simplified structure of the data.
id invest payout
[1] 1 10 -1
[2] 1 33 33
[3] 1 20 -5
[4] 2 200 33
[5] 2 33 -20
[6] 3 5 -5
[7] 3 5 -5
id is each person's id. Each person have invested many times in the sampling
period, in temporal order.
What I want to check is the correlation between invest and payout.
1. How do I run the regression for each person, with the "invest" being
devided by the mean or medium of the person's "invest"?
2. How do I plot a graph with y axis being invest/mean(invest) and x axis
being payout, all 600 people's dots superimposed on one graph?
I tried to use
for (i in 1:(dim (DF)[1]-1))
{
if (DF[i,1]=DF[i+1,1]) id.lm<- lm(invest ~ payput, data=DF)
}
But I don't know how to superimpose graphs onto each other.
Thanks a lot!
Su
[[alternative HTML version deleted]]
______________________________________________ R-help at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.