diff --git a/rubin_sim/maf/maf_contrib/periodic_star_modulation_metric.py b/rubin_sim/maf/maf_contrib/periodic_star_modulation_metric.py
index b3c18d6ba..af92e052a 100644
--- a/rubin_sim/maf/maf_contrib/periodic_star_modulation_metric.py
+++ b/rubin_sim/maf/maf_contrib/periodic_star_modulation_metric.py
@@ -187,20 +187,24 @@ def run(self, data_slice, slice_point=None):
                     noise = np.random.randn(true_lc.size) * dmag
                     # Suppress warnings about failing on covariance
                     fit_obj = PeriodicStar(t_subrun["filter"])
-                    with warnings.catch_warnings():
-                        warnings.simplefilter("ignore")
-                        # If it fails to converge,
-                        # save values that should fail later
-                        try:
-                            parm_vals, pcov = curve_fit(
-                                fit_obj,
-                                t_subrun["time"],
-                                true_lc + noise,
-                                p0=true_params,
-                                sigma=dmag,
-                            )
-                        except RuntimeError:
-                            parm_vals = true_params * 0 + np.inf
+                    # check if we have enough points
+                    if np.size(true_params) >= np.size(fit_obj):
+                        parm_vals = true_params * 0 + np.inf
+                    else:
+                        with warnings.catch_warnings():
+                            warnings.simplefilter("ignore")
+                            # If it fails to converge,
+                            # save values that should fail later
+                            try:
+                                parm_vals, pcov = curve_fit(
+                                    fit_obj,
+                                    t_subrun["time"],
+                                    true_lc + noise,
+                                    p0=true_params,
+                                    sigma=dmag,
+                                )
+                            except RuntimeError:
+                                parm_vals = true_params * 0 + np.inf
                     fits[i, :] = parm_vals
 
                 # Throw out any magnitude fits if there are no
diff --git a/rubin_sim/maf/maf_contrib/xrb_metrics.py b/rubin_sim/maf/maf_contrib/xrb_metrics.py
index 80428497f..c4786137f 100644
--- a/rubin_sim/maf/maf_contrib/xrb_metrics.py
+++ b/rubin_sim/maf/maf_contrib/xrb_metrics.py
@@ -290,7 +290,7 @@ def __init__(
         night_col="night",
         pts_needed=2,
         pts_early=2,
-        t_early=2,
+        t_early=7,
         mjd0=None,
         output_lc=False,
         badval=-666,
diff --git a/rubin_sim/maf/run_comparison/gather_summaries.py b/rubin_sim/maf/run_comparison/gather_summaries.py
index d20198246..027d61a44 100755
--- a/rubin_sim/maf/run_comparison/gather_summaries.py
+++ b/rubin_sim/maf/run_comparison/gather_summaries.py
@@ -36,9 +36,9 @@ def combine_result_dbs(run_dirs, dbfilename="resultsDb_sqlite.db"):
             run_names.append(dirname_to_runname(dname))
 
     # querry to grab all the summary stats
-    sql_q = "SELECT metrics.metric_name, metrics.metric_info_label, "
-    sql_q += "metrics.slicer_name, summarystats.summary_name, "
-    sql_q += "summarystats.summary_value "
+    sql_q = "SELECT summarystats.summary_value, "
+    sql_q += "metrics.metric_name, metrics.metric_info_label, "
+    sql_q += "metrics.slicer_name, summarystats.summary_name "
     sql_q += "FROM summarystats INNER JOIN metrics ON metrics.metric_id=summarystats.metric_id"
 
     rows = []
@@ -71,6 +71,13 @@ def combine_result_dbs(run_dirs, dbfilename="resultsDb_sqlite.db"):
             columns=col_names,
             index=[row_name],
         )
+
+        # Can have duplicate columns if MAF was run multiple times.
+        # Remove duplicates:
+        # https://stackoverflow.com/questions/14984119/
+        # python-pandas-remove-duplicate-columns
+        row = row.loc[:, ~row.columns.duplicated()].copy()
+
         rows.append(row)
     # Create final large DataFrame to hold everything
     all_cols = np.unique(np.concatenate([r.columns.values for r in rows]))
@@ -83,7 +90,8 @@ def combine_result_dbs(run_dirs, dbfilename="resultsDb_sqlite.db"):
 
     # Put each row into the final DataFrame
     for row_name, row in zip(run_names, rows):
-        result_df.loc[row_name][row.columns] = np.ravel(row.values)
+        result_df.loc[row_name, row.columns] = np.ravel(row.values)
+
     return result_df