Add dataframe output option to get_runs

CHANGELOG.md

@@ -1,5 +1,9 @@
 # Change log
 
+## v0.11.0
+
+* Support optional dataframe output from `get_runs`.
+
 ## v0.10.1
 
 * The worker process now no longer gives a long delay when a run has finished (now at most ~1 second).

setup.py

@@ -16,7 +16,7 @@
     long_description_content_type="text/markdown",
     url="https://simvue.io",
     platforms=["any"],
-    install_requires=["dill", "requests", "msgpack", "tenacity", "pyjwt", "psutil", "pydantic", "plotly"],
+    install_requires=["dill", "requests", "msgpack", "tenacity", "pandas", "pyjwt", "psutil", "pydantic", "plotly"],
     package_dir={'': '.'},
     packages=["simvue"],
     package_data={"": ["README.md"]},

simvue/__init__.py

@@ -2,4 +2,4 @@
 from simvue.client import Client
 from simvue.handler import Handler
 from simvue.models import RunInput
-__version__ = '0.10.1'
+__version__ = '0.11.0'

simvue/client.py

@@ -5,6 +5,7 @@
 
 from .serialization import Deserializer
 from .utilities import get_auth
+from .converters import to_dataframe
 
 CONCURRENT_DOWNLOADS = 10
 DOWNLOAD_CHUNK_SIZE = 8192
@@ -57,7 +58,7 @@ def get_run(self, run, system=False, tags=False, metadata=False):
         return None
 
 
-    def get_runs(self, filters, system=False, tags=False, metadata=False):
+    def get_runs(self, filters, system=False, tags=False, metadata=False, format='dict'):
         """
         Get runs
         """
@@ -73,7 +74,12 @@ def get_runs(self, filters, system=False, tags=False, metadata=False):
             return None
 
         if response.status_code == 200:
-            return response.json()
+            if format == 'dict':
+                return response.json()
+            elif format == 'dataframe':
+                return to_dataframe(response.json())
+            else:
+                return None 
 
         return None
 

simvue/converters.py

@@ -0,0 +1,35 @@
+def to_dataframe(data):
+    """
+    Convert runs to dataframe
+    """
+    import pandas as pd
+    columns = {}
+    for run in data:
+        for item in ('name', 'status', 'folder', 'created', 'started', 'ended'):
+            if item not in columns:
+                columns[item] = []
+            if item in run:
+                columns[item].append(run[item])
+            else:
+                columns[item].append()
+
+        if 'system' in run:
+            for section in run['system']:
+                if section in ('cpu', 'gpu', 'platform'):
+                    for item in run['system'][section]:
+                        if 'system.%s.%s' % (section, item) not in columns:
+                            columns['system.%s.%s' % (section, item)] = []
+                        columns['system.%s.%s' % (section, item)].append(run['system'][section][item])
+                else:
+                    if 'system.%s' % section not in columns:
+                        columns['system.%s' % section] = []
+                    columns['system.%s' % section].append(run['system'][section])
+
+        if 'metadata' in run:
+            for item in run['metadata']:
+                if 'metadata.%s' % item not in columns:
+                    columns['metadata.%s' % item] = []
+                columns['metadata.%s' % item].append(run['metadata'][item])
+
+    df = pd.DataFrame(data=columns)
+    return df