Skip to main content

Table 1 Summary of the unidirectional synthetic UMN logo inversion results

From: Using TNT-NN to unlock the fast full spatial inversion of large magnetic microscopy data sets

Method

Scenario

m (Am\(^2\))

Residual RMS (nT)

Error (\(Am^{2}\))

LH-NNLS

1

\(9.36\times 10^{-9}\)

\(3.15 \times 10^{-13}\)

\(3.51 \times \,10^{-25}\)

TNT-NN

1

\(9.36\times 10^{-9}\)

\(2.39 \times 10^{-13}\)

\(2.54 \times 10^{-25}\)

Known solution

1

\(9.36\times 10^{-9}\)

–

–

LH-NNLS

2

\(9.51\times 10^{-9}\)

0.89

\(2.56\times 10^{-12}\)

TNT-NN

2

\(9.51\times 10^{-9}\)

0.89

\(2.56\times 10^{-12}\)

Known solution

2

\(9.36\times 10^{-9}\)

–

–

LH-NNLS

3

\(9.42\times 10^{-9}\)

2.27

\(5.16\times 10^{-12}\)

TNT-NN

3

\(9.42\times 10^{-9}\)

2.27

\(5.16\times 10^{-12}\)

Known solution

3

\(9.28\times 10^{-9}\)

–

–

LH-NNLS

4

\(9.53\times 10^{-9}\)

2.37

\(5.78\times 10^{-12}\)

TNT-NN

4

\(9.53\times 10^{-9}\)

2.37

\(5.78\times 10^{-12}\)

Known solution

4

\(9.28\times 10^{-9}\)

–

–

LH-NNLS

5

\(1.07\times 10^{-8}\)

23.6

\(6.01\times 10^{-11}\)

TNT-NN

5

\(1.07\times 10^{-8}\)

23.6

\(6.01\times 10^{-11}\)

Known solution

5

\(8.51\times 10^{-9}\)

–

–

LH-NNLS

6

\(1.07\times 10^{-8}\)

23.5

\(6.01\times 10^{-11}\)

TNT-NN

6

\(1.07\times 10^{-8}\)

23.5

\(6.01\times 10^{-11}\)

Known solution

6

\(8.51\times 10^{-9}\)

–

–

  1. The first column shows the methods used to obtain the magnetic source distribution. The second column shows the scenario with which the results are associated. The third column lists the moment, m. The fourth column provides the residual RMS of the NNLS fits with the B\(_z\) field. The final column reports the error of the NNLS solution relative to the known solution