Table 1.
Precision and accuracy metrics for the dust parameters in the mock galaxy fits, for each of the test sets.
| Test | Parameter | RMS | Median Δ | Median rel. error (%) | Median fit error | MAD | Outliers (%) |
|---|---|---|---|---|---|---|---|
| Default (griz) | τV | 1.332 | 0.263 | 10.1 | 0.695 | 0.739 | 6 |
| n | 0.552 | 0.006 | 10.3 | 0.262 | 0.386 | 5.33 | |
| Default (griz+UV+NIR) | τV | 0.854 | 0.039 | 0.9 | 0.238 | 0.356 | 5.33 |
| n | 0.274 | 0.004 | 0.8 | 0.072 | 0.087 | 8 | |
| Default + Emission | τV | 0.629 | −0.014 | 0.3 | 0.280 | 0.390 | 7.33 |
| (griz+FIR) | n | 0.548 | −0.019 | 12.9 | 0.255 | 0.416 | 5.33 |
| Default (Mass+4%) | τV | 1.050 | −0.014 | 0.3 | 0.737 | 0.640 | 5.33 |
| n | 0.380 | −7.89 × 10−5 | 6.4 | 0.325 | 0.231 | 6.67 | |
| Default (Mass−4%) | τV | 0.983 | −0.019 | 0.6 | 0.720 | 0.624 | 5.33 |
| n | 0.350 | −0.004 | 3.7 | 0.291 | 0.229 | 6.67 | |
| Cardelli | τV | 1.221 | −0.036 | 1.0 | 0.789 | 0.786 | 8 |
| n | 0.840 | 0.261 | 10.6 | 1.761 | 0.720 | 4 | |
| τ = 0.1 | τV | 1.182 | 0.179 | 6.9 | 0.607 | 0.866 | 5.33 |
| n | 0.532 | −0.002 | 2.6 | 0.231 | 0.286 | 2.67 | |
| τ = 10 | τV | 0.865 | −0.061 | 1.7 | 0.715 | 0.557 | 6.67 |
| n | 0.480 | 0.027 | 12.4 | 0.324 | 0.283 | 8 | |
| ttrunc = 7.5 Gyr | τV | 1.140 | 0.042 | 3.0 | 0.239 | 0.840 | 10.67 |
| n | 0.388 | −0.027 | 4.4 | 0.280 | 0.311 | 4 | |
| ttrunc = 5 Gyr | τV | 1.038 | 0.054 | 1.4 | 0.645 | 0.537 | 8 |
| n | 0.539 | 0.031 | 9.4 | 0.220 | 0.224 | 4 | |
| ttrunc = 3 Gyr | τV | 1.038 | −2.10 × 10−4 | 0.01 | 0.570 | 0.671 | 8 |
| n | 0.512 | −0.083 | 4.6 | 0.310 | 0.307 | 6.67 | |
Notes. The median fit error was calculated based on all the error values for each of the fit parameters. As the fit error bars are asymmetric, the largest of the median values was taken in each case. The median Δ, median absolute deviation (MAD) and median relative error were calculated from the difference between the best-fit result and the “true” value for each parameter. The percentage of outliers was calculated by assuming a normal distribution for the data and identifying the points more than 2 standard deviations away from the mean.
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.