Figure 6
PD reconstructed from the difference of PC1 (\({\Delta }{\varvec{PC}}1_{i}\)), and visualization on the magnetic domain. (a) \({\Delta }{\varvec{PC}}1_{i}\) in depinning produced using Hadamard products. (a.1), (a.2), and (a.3) correspond to Up, Upper-left, and Left, respectively. (b) Visualization of total energy contribution \({\Delta }{\varvec{PC}}1_{i}\) to the original magnetic domain. The positive components with \({\Delta }{\varvec{PC}}1_{i} > 0\) and negative components with \({\Delta }{\varvec{PC}}1_{i} < 0\) are red and blue generators, respectively. (b.1), (b.2), and (b.3) correspond to Up, Upper-left, and Left, respectively. In Left, the number of generators for \({\Delta }{\varvec{PC}}1_{i} < 0\) is significantly fewer. It suggests that easier depinning is achieved. (c) Visualization of demagnetization energy contribution on the original magnetic domain using the logical sum of \({\varvec{PC}}1_{i} < 0\) and \(\hat{\user2{\beta }}^{d} > 0\). (c.1), (c.2), and (c.3) correspond to Up, Upper-left, and Left, respectively. In Left, the number of generators is fewer. The demagnetization energy barrier is significantly smaller. It indicates that the depinning is the easiest. The generators in Left do not appear near defects or at the edge of the dot. This indicates that the magnetic pole divergence and accompanying demagnetization energy loss is adequately suppressed. (d) Visualization of the exchange energy contribution to the original magnetic domain using the logical sum of \({\varvec{PC}}1_{i} < 0\) and \(\hat{\user2{\beta }}^{ex} > 0\). (d.1), (d.2), and (d.3) correspond to Up, Upper-left, and Left, respectively. The generator detects the center of the domain wall and the defects. The domain wall and defects interact, suggesting that this contributes to system stabilization.
