Thanks for replying, Dan.
I had though of running the whole thing in research but unfortunately I use the fetcher function to pull in my training data. The csv is 42MB though so it's too large to upload locally for the research platform (the limit is 25MB iirc). If there's no way to move from the backtest to research I'll just have to split the file in two and import both halves into research then concatenate the two. Can I upload and import two csv files in a single research notebook?
Also, does get_pricing in research return the exact same pricing information that data.history returns in an algorithm?