Hi again,
If I hava time series data with say for instance 1500 numeric observations reflecting price development for 1500 days on Bitcoin, variable called bitc.
Lets say I want to look at the correlation coefficient for only day 50 - 88, day 250 - 350, day 750 - 900 and so on. Meaning I do not want the correlation coefficient for the whole period as a whole, only for sub-periods.
Will I have to create new variables for the respective periods? Or may I code a command so I ask for correlation coefficient for the preferred periods?
Or which way is the easiest for a newbie. How may I code this?
Thank you in advance.
/Tor Magne
Related Posts with Time series data
Is using fixed effects and clustering acceptable for my dataset? if so: which command to use?Hi everybody, while looking for advice regarding my modeling/stata problem, I came across this forum…
eteffects vs ermHello everyone I am a sort of beginner applying causal inference and the potential outcome framewor…
Ratio of frequency of two variablesHello I have a dataset like this: Code: * Example generated by -dataex-. To install: ssc install da…
How to construct the second best response to a shock?Hi, I am using a household survey wherein I have been given money raised from top 5 sources in time…
Plotting the cumulative incidence function in the presence of competing risksHi, I am analysing a small trial where the primary outcome is time to event (recovery). All individ…
Subscribe to:
Post Comments (Atom)
0 Response to Time series data
Post a Comment