In our paper,1 we used two different dimensionless interaction constants g0 and gho. They are related to one another (see just below Eq. (25) in the original paper) by gho/g0=π/(2ρ), where ρE1/(ω) relates the infinite square well energy scale to the harmonic oscillator energy scale. We inadvertently used the wrong labels in Figs. 1–4. In these figures, all occurrences of g0 should be replaced by gho. To relate the two, use ρ = 1 in Figs. 1 and 2, and ρ = 50 remains for Figs. 3 and 4.

The matlab codes in the supplementary material are standalone. For simplicity, the code uses g0 where the harmonic oscillator is not involved (and, therefore, we have replaced ifs_only_dirac_dist.m and ifs_only_dirac_bos.m),2 reflecting Eq. (12). For the cases where the harmonic oscillator basis is used and the harmonic oscillator potential is embedded in the infinite square well potential, gho is used.

This mislabelling does not affect our qualitative results in any way but does adjust the scale of coupling strengths by a numerical factor. We wish to thank Paul Coones for alerting us to this error.

1.
MengXing
Na
and
Frank
Marsiglio
,
Am. J. Phy.
85
,
769
782
(
2017
).
2.
See the supplementary material https://www.scitation.org/doi/suppl/10.1119/5.0065753 of the original article for the updated collection of matlab routines used to perform the calculations described in this paper.

Supplementary Material