O3-review/GW150914_IMRPhenomPv2_lalinference.md

@@ -56,3 +56,23 @@ Run settings are same as in regular bilby run, but the workflow is executed slig
 The ini file does not include the sampler args. These are passed to parallel_bilby_analysis within the slurm submit file:
 
 `mpirun parallel_bilby_analysis --nlive 1500`
+
+## How we get data
+
+The first step in the pipeline `python get_gwf.py` uses `gwpy` to get strain data in gwf format. The `.gwf` file is then passed into the subsequent stages of the pipeline. An example of how the data retrieval works is shown below:
+
+```
+
+from gwpy.timeseries import TimeSeries
+
+duration = 4
+trigger_time = 1126259462.3910
+end_time = trigger_time +2
+start_time = end_time - duration
+
+dataH1 = TimeSeries.get('H1:DCS-CALIB_STRAIN_C02', start_time, end_time)
+dataL1 = TimeSeries.get('L1:DCS-CALIB_STRAIN_C02', start_time, end_time)
+
+TimeSeries.write(dataH1, target='dataH1.gwf',format='gwf')
+TimeSeries.write(dataL1, target='dataL1.gwf',format='gwf')
+```
\ No newline at end of file