From 610a20c5f8dde566741119648f2d93dec1df7df4 Mon Sep 17 00:00:00 2001
From: Shashank <samshashcool@gmail.com>
Date: Sun, 17 Feb 2019 17:48:12 +0530
Subject: [PATCH 1/2] Adding Documentation example for combining DocString

---
 arviz/data/inference_data.py | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/arviz/data/inference_data.py b/arviz/data/inference_data.py
index 8bb72d868d..f043f9b69b 100644
--- a/arviz/data/inference_data.py
+++ b/arviz/data/inference_data.py
@@ -91,7 +91,14 @@ def to_netcdf(self, filename, compress=True):
         return filename
 
     def __add__(self, other):
-        """Concatenate two InferenceData objects."""
+        """Concatenate two InferenceData objects.
+        example
+        --------
+        A._groups == ["posterior", "posterior_predictive"]
+        B._groups == ["prior", "prior_predictive"]
+        C=A+B
+        C._groups ==["posterior", "posterior_predictive", "prior", "prior_predictive"]
+        """
         return concat(self, other, copy=True, inplace=False)
 
 
@@ -116,6 +123,19 @@ def concat(*args, copy=True, inplace=False):
     InferenceData
         A new InferenceData object by default.
         When `inplace==True` merge args to first arg and return `None`
+    
+    example
+    -------
+    A._groups == ["posterior", "posterior_predictive"]
+    B._groups == ["prior", "prior_predictive"]
+    C = az.concat(A, B)
+    C._groups ==["posterior", "posterior_predictive", "prior", "prior_predictive"]
+    
+    When inplace=True
+    -----------------
+    az.concat(A, B, inplace=True)
+    A._groups ==["posterior", "posterior_predictive", "prior", "prior_predictive"]
+    
     """
     if len(args) == 0:
         return InferenceData()

From 3949bee46bcac62783ba631b8e15f45d5eb4b7ec Mon Sep 17 00:00:00 2001
From: Shashank <samshashcool@gmail.com>
Date: Sun, 17 Feb 2019 19:35:23 +0530
Subject: [PATCH 2/2] Adding the test cases and solving the issue of creation
 of empty file

---
 arviz/data/inference_data.py | 24 +++++++++++++++---------
 arviz/tests/test_data.py     | 16 ++++++++++++++--
 2 files changed, 29 insertions(+), 11 deletions(-)

diff --git a/arviz/data/inference_data.py b/arviz/data/inference_data.py
index f043f9b69b..a756bcb3b9 100644
--- a/arviz/data/inference_data.py
+++ b/arviz/data/inference_data.py
@@ -80,15 +80,21 @@ def to_netcdf(self, filename, compress=True):
             Location of netcdf file
         """
         mode = "w"  # overwrite first, then append
-        for group in self._groups:
-            data = getattr(self, group)
-            kwargs = {}
-            if compress:
-                kwargs["encoding"] = {var_name: {"zlib": True} for var_name in data.variables}
-            data.to_netcdf(filename, mode=mode, group=group, **kwargs)
-            data.close()
-            mode = "a"
-        return filename
+	#if the netcdf file previously exists the append the file 
+        if self._groups:
+            for group in self._groups:
+                data = getattr(self, group)
+                kwargs = {}
+                if compress:
+                    kwargs["encoding"] = {var_name: {"zlib": True} for var_name in data.variables}
+                data.to_netcdf(filename, mode=mode, group=group, **kwargs)
+                data.close()
+                mode = "a"
+            return filename
+        else: # else if the file doesnot exists previously creating the empty netcdf file
+            empty_file=nc.Dataset(filename,mode="w",format="NETCDF4")
+            empty_file.close()
+            return filename
 
     def __add__(self, other):
         """Concatenate two InferenceData objects.
diff --git a/arviz/tests/test_data.py b/arviz/tests/test_data.py
index d65d6ca1b3..7e75839978 100644
--- a/arviz/tests/test_data.py
+++ b/arviz/tests/test_data.py
@@ -672,7 +672,7 @@ def test_io_function(self, data, eight_schools_params):
         assert hasattr(inference_data3, "posterior")
         os.remove(filepath)
         assert not os.path.exists(filepath)
-
+     
     def test_io_method(self, data, eight_schools_params):
         inference_data = self.get_inference_data(  # pylint: disable=W0612
             data, eight_schools_params
@@ -690,7 +690,19 @@ def test_io_method(self, data, eight_schools_params):
         assert hasattr(inference_data2, "posterior")
         os.remove(filepath)
         assert not os.path.exists(filepath)
-
+    
+    #New test case added for the empty file to be created
+    def test_new_file(self):
+        inference_data=InferenceData()
+        here=os.path.dirname(os.path.abspath(__file__))
+        data_directory=os.path.join(here,"saved_models")
+        filepath=os.path.join(data_directory,"io_new_file.nc")
+        assert not os.path.exists(filepath)
+        to_netcdf(inference_data,filepath)
+        assert os.path.exists(filepath)
+        assert os.path.getsize(filepath)>0
+        os.remove(filepath)
+        assert not os.path.exists(filepath)
 
 class TestPyMC3NetCDFUtils:
     @pytest.fixture(scope="class")