Table 5
Comparison of bias, fraction of outliers, and variance between photo-z computed with PICZL for various subsamples.
| Code | Sample | Number of sources | Bias | Variance | Fraction of outliers |
|---|---|---|---|---|---|
| 〈∆ɀ〉 | σNMAD | η [%] | |||
| PICZL | Validation | 8098 | −0.004 | 0.045 | 5.6 |
| Type: PSF | 6083 | −0.001 | 0.049 | 6.5 | |
| Type: EXT | 2015 | −0.011 | 0.031 | 3.3 | |
| S/N > 3 | 6844 | −0.002 | 0.044 | 5.4 | |
| S/N < 3 | 1254 | −0.009 | 0.042 | 7.6 | |
| w/ i-band | 7127 | −0.002 | 0.044 | 5.6 | |
| w/o i-band | 971 | −0.011 | 0.044 | 7.1 | |
| optically selected AGN | 4636 | −0.002 | 0.040 | 3.9 | |
| X-ray selected AGN | 3462 | −0.005 | 0.050 | 8.9 | |
| isolated AGN | 3934 | 0.005 | 0.042 | 5.5 | |
| w/faint neighbour | 407 | 0.004 | 0.051 | 6.4 | |
| w/bright neighbour | 109 | −0.021 | 0.081 | 20.1 | |
| CSC2 | 416 | −0.007 | 0.046 | 8.3 | |
| CIRCLEZ | CSC2 | 416 | −0.009 | 0.055 | 12.3 |
Notes. Columns: (1)All sources; (2)split by type (PSF vs. EXT); (3)split by signal-to-noise ratio (S/N) of all available bands; (4)split by availability of the i-band; (5)split by selection criteria; (6)split by the presence of a faint [(mAGN − mNeigh.) within −3 ≤ ∆mag < −1 for all available bands] or bright [(mAGN − mNeigh.) within Amag > −1 for all available bands] neighbor within a five arcsecond radius. Additionally, results derived for the CSC2 blind sample for both PICZL and CIRCLEZ are provided for comparison.
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.