gmwm

gmwm copied to clipboard

**Feature:** Right now, we use a non-descript approach to generate CIs for AVAR & HVAR. Thus, we need to find the appropriate way to create the CIs. Potential ideas: -...

coatless

Estimation stability

10

Apparently the GMWM (gmwm.imu()) slighlty changes estimates for the same dataset when changing R sessions. I guess it's due to the guessing algorithm but it would be good to find...

robertomolinari

This is how it will be implemented  $$Z_t^(1) = X_t + Y_t^(1)$$ $Z_t^(2) = X_t + Y_t^(2)$ $Z_t^(1)$ and $Z_t^(2)$ are the stock prices of the same company but...

Wenchao-Yang

**Issue:** Currently, the user can either supply the starting values or have the program begin guessing all the values. Sometimes it would be helpful to fix certain parameters and have...

coatless

To estimate the parameters of stationary latent time series models with relatively small sample size (e.g. T < 1000) we need to implement the MLE of such models. This can...

stephaneguerrier

**Issue:** Right now, we only have comparison tools to highlight the difference of the wavelet variances. Given that we are adding more robust functions (avar, hadamard, and so on), it...

coatless

Need to improve the vignettes that ship with the package. - Finish computational object section of the manipulating GMWM objects - Give SSM examples

coatless

**Issue:** More helpful to see the top 4 models per process than what each best model for each process is. **Solution:** Create a wrapper function to call `compare.models()` with the...

coatless

**Issue** The `rank.models` and the `auto.imu` functions release only the best models `gmwm` object. However, there are _p_ more `gmwm` objects created internal but never surfaced. **Solution** 1. Verify that...

coatless

coatless