Redshift and wavenumber dependence of the relative error in emulating the best-fitting power spectrum. The upper panels shows the dependence on wavenumber (for fixed redshift) and the lower panels depict the redshift dependence (for fixed wavenumber). For the left-most panels, we emulate the power spectrum using the true parameters from the test data set. For the panels in the second column from the left, we emulate the power spectrum using the best-fitting parameters derived from the network of ideal data set. For the panels in the third column from the left, we use the best-fitting parameters derived from the network of mock SKA data, but for the error we measure the prediction of the real power spectrum, i.e. we apply the emulator trained without SKA noise. For the panels in the right-most column, we use the best-fitting parameters derived from the network of SKA thermal noise case and otherwise do the same as for the third column. Note that the plots in this figure show all 25 z values and 32 k values.