Simulated Data for Completeness and Reliability

The completeness and reliability of the Kepler pipeline must be quantified in order to calculate robust planet occurrence rates. The Kepler project generated multiple kinds of simulated data to challenge the pipeline and examine the results; these data are served below to facilitate generation of custom completeness and reliability calculations. For the Kepler project completeness and reliability calculations, see the Kepler Completeness and Reliability page.

Skip to:

• Modified Light Curve Products
• Flux-level Transit Injection Products

Modified Light Curve Products

Five sets of simulated data were created during the production of the Q1-Q17 DR25 KOI catalog:

• Injected Group 1 (INJ1) contains a single on-target injection for 146,294 observed targets occupying 64 channels in the Kepler focal plane. Each injection mimics a planet transiting the specified target.
• Injected Group 2 (INJ2) contains a single off-target injection for 33,978 observed targets occupying 16 channels in the Kepler focal plane. Each injection mimics a foreground or background transiting planet or eclipsing binary along the line of sight to the specified target.
• Injected Group 3 (INJ3) contains two on-target injections for 9,856 observed targets occupying 4 channels in the Kepler focal plane. Each injection pair has a common period so as to simulate an eclipsing binary on the specified target.
• Inverted Group (INV) contains 198,640 targets that have had their observed light curves inverted immediately before the planet search to produce a set of TCEs that are likely to be false positives.
• Scrambled Group (SCR1, SCR2, & SCR3) contains 198,640 targets that have had their observed light curves scrambled by rearranging the data in yearly chunks immediately before the planet search to produce a set of TCEs that are likely to be false positives. Refer to KSCI-19114-002.pdf for the order in which the quarters have been scrambled in each group.

The relevant data products are listed in the table below for each group:

• Light curves with injected transits: The signals were injected at the pixel level (i.e., into the CAL output) and the normal photometric pipeline (i.e., modules PA and PDC) was run to create standard light curve products.
• DV summaries and DV reports: A normal search was conducted on the injected light curves with pipeline modules TPS and DV to produce standard DV summaries and reports for each TCE identified.
• Average Detection Efficiency: The results of the full pipeline run on the injected light curves (i.e., PA, PDC, TPS, DV) can be used to compute the average detection efficiency of the SOC 9.3 pipeline for an arbitrary subset of targets.
• TCERT Reports: The DR25 Robovetter metrics were produced for all simulated (and observed) data sets and these standard reports were generated to allow manual evaluation of any simulated (or observed) TCE.
• Robovetter Results: The DR25 Robovetter was run and the dispositions and flags preserved for all simulated TCEs for use in determining completeness and effectiveness of the dispositioning process.
• Robovetter input/output: The Robovetter metrics are formatted as input files and the Robovetter outputs are preserved for all simulated data sets.

† The file size refers to the destination data set, which requires an initial download of a wget script that is much smaller (~15 MB). For more information, see the wget help page. For best performance, do not run more than 4 wget scripts in parallel. Running additional parallel scripts will not decrease the total download duration.

‡ The Robovetter results, Robovetter input, and Robovetter output files for the observed TCEs are available in the Q1-Q17 DR25 section of the TCE documentation page.

Flux-level Transit Injection Products

The per-target detection efficiency was generated by injecting thousands of signals into Kepler light curves. These files are divided into five runs, as described in the documentation (KSCI-19109-002.pdf).

Last updated: 19 December 2017