Abstract: The IceCube Neutrino Observatory detects neutrinos by instrumenting 1 km3 of deep glacial ice with photomultiplier tubes. These detections allow for the study of possible astrophysical neutrino sources and the measurement of the diffuse astrophysical neutrino flux. The flavor composition of the measured flux is a very important component in understanding and modeling the astrophysical neutrino flux. In addition, an improved measurement of the flavor composition of the astrophysical neutrino flux at Earth allows us to infer the flavor composition at the sources and thus provides insights into neutrino production mechanisms. I will present the plan for a new flavor analysis using a combination of datasets used within the IceCube Collaboration. Previous flavor measurements with IceCube focused on individual samples targeting distinct event types, mostly dominated by events with interaction vertices inside the detector. Building on existing efforts, I plan to perform such a measurement on an expanded dataset. Such a sample includes contained and uncontained events, incorporates improved ice modeling, and achieves the highest statistics for the high-energy neutrino flux. These updates in event classification promise improvement in the sensitivity of the flavor measurement. Identifying double-cascades among tracks and cascades is a continuous effort, and the main focus of my talk.