Injection example eccentric_inspiral.py unbale to recover the parameters

[CalNaTNS]

Hello everyone,
I recently tried the example eccentric_inspiral.py and found that the result didn't recover the injected parameters, especially the eccentricity, which deviated greatly from the injected value e=0.1.

fig.1 The posterior for eccentricFD and GW150914
The following are my versions of bilby and other related packages:

astropy                   6.0.1            py39h44dd56e_0    conda-forge
astropy-iers-data         0.2024.7.15.0.31.42    pyhd8ed1ab_0    conda-forge
bilby                     2.3.0              pyhd8ed1ab_0    conda-forge
bilby.cython              0.5.1            py39hf15bc26_0    conda-forge
bilby_pipe                1.3.2              pyhd8ed1ab_0    conda-forge
dynesty                   2.1.4              pyhd8ed1ab_0    conda-forge
gwosc                     0.7.1              pyhd8ed1ab_0    conda-forge
gwpy                      3.0.9              pyhd8ed1ab_0    conda-forge
lal                       7.1.7           fftw_py39h90429df_100    conda-forge
lalapps                   8.1.1            py39ha640902_0    conda-forge
lalburst                  1.5.12           py39hb9d737c_0    conda-forge
lalframe                  2.0.0            py39hb9d737c_1    conda-forge
lalinference              3.0.3           nompi_py39h9f32d32_100    conda-forge
lalinference-data         3.0.3                ha770c72_0    conda-forge
lalinspiral               2.0.6            py39hb9d737c_0    conda-forge
lalmetaio                 2.1.2            py39hb9d737c_1    conda-forge
lalpulsar                 4.0.0            py39hb9d737c_2    conda-forge
lalpulsar-data            4.0.0                ha770c72_2    conda-forge
lalsimulation-data        3.1.2                ha770c72_1    conda-forge
lalsuite                  7.11                     pypi_0    pypi
numpy                     1.26.4           py39h474f0d3_0    conda-forge
parallel-bilby            2.0.2              pyhd8ed1ab_0    conda-forge
pycbc                     2.4.2            py39hebabecc_0    conda-forge
pymultinest               2.12               pyh707e725_0    conda-forge
python                    3.9.19          h0755675_0_cpython    conda-forge
python-lal                7.1.7           fftw_py39h4a6fd98_100    conda-forge
python-lalburst           1.5.12           py39h78aa3b4_0    conda-forge
python-lalframe           2.0.0            py39hd257fcd_1    conda-forge
python-lalinference       3.0.3            py39hf5a7330_0    conda-forge
python-lalinspiral        2.0.6            py39hd257fcd_0    conda-forge
python-lalmetaio          2.1.2            py39hd257fcd_1    conda-forge
python-lalpulsar          4.0.0            py39hd257fcd_2    conda-forge
python-lalsimulation      3.1.2            py39hd257fcd_1    conda-forge
scipy                     1.13.1           py39haf93ffa_0    conda-forge

Besides, I tried multiple spin-aligned waveforms like IMRPhenomXHM, and SEOBNRv4HM_ROM, and all returned bad results. Therefore I suppose it is not a specific problem with EccentricFD. Are any settings important for spin-aligned waveform PE missing here that resulted in the phenomena?

Thanks!
Honglue