Table 2.
Parameter survey results for April 21 band 3 data.
| DIFMAP (1260 Param. combinations; 303 in Top Set) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Compact | 0.4 | 0.5 | 0.6 | 0.7 | 0.8 | |||
| flux (Jy) |  |
17% | 24% | 23% | 23% | |||
| ALMA | 0.01 | 0.1 | 0.3 | 0.5 | 0.7 | 1.0 | ||
| weight factor | 20% |  |
26% | 17% | 10% | 3% | ||
| Mask diam. | 40 | 50 | 60 | 70 | 80 | 90 | 100 | |
| (μas) | 0% | 4% |  |
27% | 23% | 12% | 7% | |
| (u, v) weight | –2 | –1 | 0 | |||||
| exponent κ | 32% |  |
9% | |||||
| Stop | Flux reached | Δrms ≤ 10−4 | ||||||
| condition |  |
48% | ||||||
| eht-imaging (12288 Param. combinations; 874 in Top Set) | ||||||||
| Compact | 0.4 | 0.5 | 0.6 | 0.7 | ||||
| flux (Jy) | 18% |  |
29% | 26% | ||||
| Init./MEM | 40 | 50 | 60 | |||||
| FWHM (μas) |  |
39% | 4% | |||||
| Systematic | 0% | 1% | 2% | 5% | ||||
| error | 14% |  |
30% | 34% | ||||
| Regularizer: | 0 | 1 | 10 | 100 | ||||
| MEM | 0% | 0% | 5% |  |
||||
| TV | 33% |  |
30% | 0% | ||||
| TSV | 31% |  |
35% | 1% | ||||
| ℓ1 | 26% |  |
25% | 24% | ||||
| SMILI (36288 Param. combinations; 5333 in Top Set) | ||||||||
| Compact | 0.4 | 0.5 | 0.6 | 0.7 | 0.8 | 0.9 | 1.0 | |
| flux (Jy) | 10% | 18% |  |
18% | 15% | 13% | 10% | |
 Soft mask |
40 | 50 | 60 | 70 | 80 | 90 | 100 | 110 |
| FWHM (μas) | 13% | 15% | 15% |  |
12% | 11% | 11% | 11% |
| Systematic | 0% | 1% | 2% | |||||
| error |  |
33% | 30% | |||||
| Regularizer: | 0 | 10 | 102 | 103 | 104 | 105 | ||
| TV |  |
19% | 20% | 23% | 19% | 0% | ||
| TSV |  |
19% | 20% | 23% | 18% | 0% | ||
| 0 | 10−2 | 10−1 | 1 | 10 | 102 | |||
 |
0% | 0% | 4% | 35% |  |
2% | ||
Notes. Below each parameter values we specify the fraction of the Top Set parameter combinations that include that value. Boxed parameters are those corresponding to the fiducial images. The fiducial parameters are determined by identifying the parameter combinations that jointly perform best on all the synthetic data sets; the fiducial parameters do not necessarily correspond with the parameters that have the largest share in the Top Set. Regularization terms are maximum entropy method (MEM), total variation (TV), total squared variation (TSV), the ℓ1 norm (ℓ1), and the weighted-ℓ1 (
), and these are described in M 87∗ 2017 IV.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.