v0.10.0.md

@@ -12,7 +12,7 @@
 
 [pesummary_sql_review_notebook.html](uploads/280f1780780d3aa5cf86e0e879e8b8e1/pesummary_sql_review_notebook.html)
 
-## Reviewer Comments:
+## Reviewer comments:
 
 02/10/2020 13:26 @philip.relton:
 - I have written and read a SQL database from a set of samples from a Bilby run
@@ -20,6 +20,28 @@
 - The differences are just due to the difference in floating point storage between
   SQL and JSON. I believe this difference is small enough to be negligible
 
+# CSV file format
+
+## Related MRs
+
+* https://git.ligo.org/lscsoft/pesummary/-/merge_requests/466
+
+## Test
+
+To check that posterior samples can both be written to CSV and read back in using pesummary's `io` module, we ran the following test:
+
+```python
+from pesummary.io import read
+
+f = read()
+f.write(file_format="csv", filename=".csv", outdir="./")
+g = read("./.csv")
+assert f.parameters == g.parameters
+assert f.samples == g.samples
+```
+
+## Reviewer comments
+
 # Preliminary watermark
 
 ## Related MRs
@@ -36,7 +58,7 @@ To ensure the Preliminary watermark is added when required we generated 4 summar
 * [Double analysis with no PSD and no config provided](https://ldas-jobs.ligo.caltech.edu/~charlie.hoy/projects/pesummary_review/v0.10.0/watermark/double_no_config_no_psd/home.html): here we ran pesummary with 2 results files and did not provide a config or psd for either case. We expect the Preliminary watermark to be displayed on all html pages
 * [Double analysis, one ready one not](https://ldas-jobs.ligo.caltech.edu/~charlie.hoy/projects/pesummary_review/v0.10.0/watermark/double_one_ready_one_not/home.html): here we ran pesummary with 2 result files. For one analysis, we provided the PSD and configuration file. For the other, we only provided the configuration file. We therefore expect that the Preliminary watermark is only added to the `no_psd` pages as only that run is not reproducible.
 
-## Reviewer Comments
+## Reviewer comments
 
 # Fetch GWTC-2 result files
 
@@ -60,7 +82,7 @@ from pesummary.gw.fetch import fetch_open_data
 from pesummary.io import read
 
 for event in ["GW190412", "GW190521"]:
-    f = fetch_open_data(event, path=f"{event}.h5")
+    f = fetch_open_data(event, unpack=True, read_file=True, path=f"{event}.h5")
     g = read(f"{event}/{event}.h5")
     assert sorted(f.labels) == sorted(g.labels)
     np.testing.assert_almost_equal(f.samples, g.samples)