-
Notifications
You must be signed in to change notification settings - Fork 14
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Refactor design/response into Python #28
Comments
Following up on discussion with @kostyat @nickdeveaux today (just noting this here before I forget; we can discuss in person if that's easier): is there any reason not to use a downsampling method of an irregular time series into a regular one to take care of the "chains of small |
I don't think downsampling the time series would be a good way to go. It could be something we might want to do in the future, but since our current goal is to implement the same algorithm as we have in R into Python, it's not a good idea because it will result in very different outputs. It will mean that the number of samples will change, and for longer time series with many time points we will end up using only some of those time points, and thus such time series will be given less weight than identical time series with less frequent measurements, even though we have more data for them. I hope that explanation makes sense, but I can explain further if not. |
@kostyat Got it, makes sense. I'm currently working on building the tree/digraph representation of the conditioned time series measurements that we discussed. Could either of you (@kostyat @nickdeveaux) describe what the |
Relatedly, I had some fun visualizing the branching of the b. subtilis time series with |
Hi all, creating this issue to track progress on this work.
The goal is to translate
R_code/design-and_response.R
into Python. Some progress is in thedesign_response_refactor
branch of my fork, here:https://github.com/e5c/inferelator_ng/blob/design_response_refactor/inferelator_ng/design_response.py
I'm hoping to get an in-person explanation of the steps handling the time series (starting at https://github.com/simonsfoundation/inferelator_ng/blob/master/inferelator_ng/R_code/design_and_response.R#L409) as I'm guessing that explanation will be more translatable to python/pandas than the R code.
The text was updated successfully, but these errors were encountered: