Merge pull request #139 from NumCosmo/improve_test_numcosmo

Improve numcosmo command line tool

numcosmo_py/external/cosmosis.py

@@ -26,7 +26,7 @@
 
 import os
 import math
-from typing import List, Dict, Tuple, Optional, Iterator
+from typing import List, Dict, Tuple, Optional
 from pathlib import Path
 from enum import Enum
 
@@ -62,7 +62,7 @@ class NonLinearMatterPowerSpectrum(str, Enum):
     HALOFIT = "halofit"
 
 
-def convert_parameter(p: Parameter, required_parameters: Iterator[str]) -> Ncm.SParam:
+def convert_parameter(p: Parameter, required_parameters: List[str]) -> Ncm.SParam:
     """Converts a Cosmosis parameter to a NumCosmo parameter."""
 
     matched_name = next(
@@ -109,7 +109,7 @@ def convert_parameter(p: Parameter, required_parameters: Iterator[str]) -> Ncm.S
 
 def convert_single_model(
     sampling_parameters_section: str,
-    required_parameters: Iterator[str],
+    required_parameters: List[str],
     model_name: str,
     parameters: List[Parameter],
 ) -> Tuple[Ncm.ModelBuilder, Ncm.Model]:
@@ -149,7 +149,7 @@ def convert_single_model(
 def convert_models(
     sampling_parameters_sections: List[str],
     model_names_list: List[str],
-    required_parameters: Iterator[str],
+    required_parameters: List[str],
     parameters: List[Parameter],
     model_builders: Ncm.ObjDictStr,
     mset: Ncm.MSet,
@@ -253,7 +253,9 @@ def convert_single_likelihood(
     likelihood.peek_dataset().append_data(firecrown_data)
 
     firecrown_likelihood = numcosmo_factory.get_firecrown_likelihood()
-    required_parameters = firecrown_likelihood.required_parameters().get_params_names()
+    required_parameters = list(
+        firecrown_likelihood.required_parameters().get_params_names()
+    )
 
     # Converts the sampling parameters sections to NumCosmo models.
     convert_models(

tools/numcosmo

@@ -312,16 +312,18 @@ class RunFit(RunCommonOptions):
                 " than the given tolerance (abstol, reltol)."
             ),
         ),
-    ] = None
+    ] = (None, None)
 
     def __post_init__(self):
         super().__post_init__()
         self.fit.log_info()
 
-        if self.restart is None:
+        abstol, reltol = self.restart
+
+        if abstol is None or reltol is None:
             self.fit.run(self.run_messages.genum)
         else:
-            if self.restart[0] <= 0.0 and self.restart[1] <= 0.0:
+            if abstol <= 0.0 and reltol <= 0.0:
                 raise RuntimeError(f"Invalid tolerance for restart {self.restart}.")
             output_filename = (
                 None
@@ -330,8 +332,8 @@ class RunFit(RunCommonOptions):
             )
             self.fit.run_restart(
                 self.run_messages.genum,
-                self.restart[0],
-                self.restart[1],
+                abstol,
+                reltol,
                 None,
                 output_filename,
             )