The datasets are part of a project exploring the onset of phase transitions, the development of microstrain and the lattice parameters during fast compression rates, resembling those during e.g. propagation seismic shockwaves and large body impacts. The strain rate conditions achieved in our experiments are between those reached in the common static (DAC) and shock experiments. The current datasets contain a selected set of xy.files of azimuthal integrated 2D-diffraction images of either Mg0.2Fe0.8O sample or Fe (Iron) - with and without Pt (platinum) pressure marker and with and without Ne (Neon) pressure transmitting medium (PTM). Diffraction images were collected at ambient temperature (in 2019 and 2020) at the ECB P02.2 Beamline, PETRA III, DESY, Germany using the piezo-driven dynamic diamond anvil cells (dDAC) in combination with fast LAMBDA GaAs 2M detectors at 25.6 keV (0.4843 Å). The datasets are provided as test datasets for machine learning application on spectra classification. The version will be further updated once the whole work has been published and the full datasets can be made available. Please find below the link towards the machine learning application. https://github.com/European-XFEL-examples/panosc-ml-spectra-classification