Merge branch 'master' of https://github.com/ml4ai/delphi

Author: adarshp

apt.txt

@@ -1,3 +1,4 @@
+default-jre
 python3-dev
 graphviz
 libgraphviz-dev

data/postBuild

@@ -0,0 +1,7 @@
+jupyter contrib nbextension install --user
+jupyter nbextension enable --py widgetsnbextension
+jupyter nbextension enable toc2/main
+
+# Notebooks w/ extensions that auto-run code must be "trusted" to work the first
+# time
+jupyter trust notebooks/*.ipynb

delphi/AnalysisGraph.py

@@ -114,7 +114,7 @@ def infer_transition_model(
         if adjective_data is None:
             adjective_data = adjectiveData
 
-        gb = pd.read_csv(adjectiveData, delim_whitespace=True).groupby(
+        gb = pd.read_csv(adjective_data, delim_whitespace=True).groupby(
             "adjective"
         )
 

delphi/assembly.py

@@ -1,6 +1,6 @@
 from datetime import datetime
 from delphi.paths import concept_to_indicator_mapping, data_dir
-from .utils import exists, flatMap, flatten
+from .utils import exists, flatMap, flatten, get_data_from_url
 from .random_variables import Delta, Indicator
 from typing import *
 from indra.statements import Influence, Concept
@@ -229,8 +229,9 @@ def get_data(filename: str) -> pd.DataFrame:
     return df
 
 
-def get_mean_precipitation(year: int, cycles_output=data_dir + "/weather.dat"):
-    df = pd.read_table(cycles_output)
+def get_mean_precipitation(year: int):
+    url="http://vision.cs.arizona.edu/adarsh/export/demos/data/weather.dat"
+    df = pd.read_table(get_data_from_url(url))
     df.columns = df.columns.str.strip()
     df.columns = [c + f" ({df.iloc[0][c].strip()})" for c in df.columns]
     df.drop([0], axis=0, inplace=True)
@@ -278,11 +279,12 @@ def process_variable_name(x: str):
     return " ".join(xs[0:2])
 
 
-def construct_concept_to_indicator_mapping(n: int = 2) -> Dict[str, List[str]]:
+def construct_concept_to_indicator_mapping(n: int = 2,
+        mapping=concept_to_indicator_mapping) -> Dict[str, List[str]]:
     """ Create a dictionary mapping high-level concepts to low-level indicators """
 
     df = pd.read_table(
-        concept_to_indicator_mapping,
+        mapping,
         usecols=[1, 3, 4],
         names=["Concept Grounding", "Indicator Grounding", "Score"],
     )

delphi/export.py

@@ -10,6 +10,16 @@
 import pickle
 from .execution import construct_default_initial_state
 from datetime import datetime
+import platform
+
+operating_system = platform.system()
+
+if operating_system == "Darwin":
+    font = "Gill Sans"
+elif operating_system == "Windows":
+    font = "Candara"
+else:
+    font = "Ubuntu"
 
 # ==========================================================================
 # Export
@@ -33,7 +43,7 @@ def to_agraph(G, *args, **kwargs) -> AGraph:
             "dpi": 227,
             "fontsize": 20,
             "rankdir": kwargs.get("rankdir", "TB"),
-            "fontname": "Gill Sans",
+            "fontname": font,
         }
     )
 
@@ -42,7 +52,7 @@ def to_agraph(G, *args, **kwargs) -> AGraph:
             "shape": "rectangle",
             "color": "#650021",
             "style": "rounded",
-            "fontname": "Gill Sans",
+            "fontname": font,
         }
     )
 

delphi/parameterization.py

@@ -7,23 +7,17 @@
 
 
 def parameterize(
-    G: AnalysisGraph, time: datetime, data: Optional[pd.DataFrame] = None
+        G: AnalysisGraph, time: datetime, data = south_sudan_data
 ) -> AnalysisGraph:
     """ Parameterize the analysis graph.
 
     Args:
         G
         time
-        data
+        datafile
     """
 
-    if data is not None:
-        G.data = data
-    else:
-        if G.data is None:
-            G.data = get_data(south_sudan_data)
-        else:
-            pass
+    G.data = get_data(data)
 
     nodes_with_indicators = [
         n for n in G.nodes(data=True) if n[1]["indicators"] is not None

delphi/program_analysis/scopes.py

@@ -82,7 +82,7 @@ def to_agraph(self):
         elif operating_system == "Windows":
             font = "Consolas"
         else:
-            font = "Courier"
+            font = "Ubuntu Mono"
 
         A.node_attr["fontname"] = font
         A.graph_attr["fontname"] = font

delphi/quantification.py

@@ -8,7 +8,7 @@
 
 
 def map_concepts_to_indicators(
-    G: AnalysisGraph, n: int = 1, manual_mapping: Optional[dict] = None
+        G: AnalysisGraph, n: int = 1, mapping_file: Optional[str] = None
 ) -> AnalysisGraph:
     """ Add indicators to the analysis graph.
 
@@ -17,14 +17,11 @@ def map_concepts_to_indicators(
         n
         manual_mapping
     """
-    mapping = construct_concept_to_indicator_mapping(n=n)
+    mapping = construct_concept_to_indicator_mapping(n, mapping_file)
 
     for n in G.nodes(data=True):
         n[1]["indicators"] = get_indicators(
             n[0].lower().replace(" ", "_"), mapping
         )
-        if manual_mapping is not None:
-            if n[0] in manual_mapping:
-                n[1]["indicators"] = manual_mapping[n[0]]
 
     return G

delphi/utils.py

@@ -281,6 +281,10 @@ def download_file(url: str, filename: str):
         request.urlretrieve(url, filename, reporthook)
 
 
+def get_data_from_url(url: str):
+    return request.urlopen(url)
+
+
 def _change_directory(destination_directory):
     cwd = os.getcwd()
     os.chdir(destination_directory)

notebooks/PI Meeting 2018 Demo.ipynb

@@ -65,11 +65,10 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import urllib.request\n",
+    "from delphi.utils import get_data_from_url\n",
     "import pickle\n",
     "url = 'http://vision.cs.arizona.edu/adarsh/export/demos/data/pi_mtg_demo.pkl'\n",
-    "with urllib.request.urlopen(url) as f:\n",
-    "    sts = pickle.load(f) "
+    "sts = pickle.load(get_data_from_url(url)) "
    ]
   },
   {
@@ -195,6 +194,16 @@
     "## Mapping concepts to indicators"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%load_ext autoreload\n",
+    "%autoreload 2"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -206,7 +215,9 @@
    "outputs": [],
    "source": [
     "from delphi.quantification import map_concepts_to_indicators\n",
-    "map_concepts_to_indicators(G, 2)\n",
+    "url = 'http://vision.cs.arizona.edu/adarsh/export/demos/data/concept_to_indicator_mapping.txt'\n",
+    "mapping = get_data_from_url(url)\n",
+    "map_concepts_to_indicators(G, 2, mapping)\n",
     "visualize(G, indicators=True)"
    ]
   },
@@ -233,6 +244,7 @@
    "cell_type": "code",
    "execution_count": null,
    "metadata": {
+    "scrolled": false,
     "slideshow": {
      "slide_type": "subslide"
     }
@@ -242,7 +254,8 @@
     "from datetime import datetime\n",
     "from delphi.parameterization import parameterize\n",
     "date = datetime(2014, 1, 1)\n",
-    "parameterize(G, datetime(2014, 1,1))\n",
+    "url = 'http://vision.cs.arizona.edu/adarsh/export/demos/data/south_sudan_data.csv'\n",
+    "parameterize(G, datetime(2014, 1,1), get_data_from_url(url))\n",
     "visualize(G, indicators=True, indicator_values = True,\n",
     "          graph_label=f'Causal Analysis Graph for South Sudan, {date.year}')"
    ]
@@ -258,15 +271,6 @@
     "## Infer transition model"
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "jt.display_image('images/delphi_model.png')"
-   ]
-  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -277,7 +281,8 @@
    },
    "outputs": [],
    "source": [
-    "%time G.infer_transition_model(res=10000)"
+    "url = 'http://vision.cs.arizona.edu/adarsh/export/demos/data/adjectiveData.tsv'\n",
+    "G.infer_transition_model(get_data_from_url(url), 10000)"
    ]
   },
   {
@@ -317,16 +322,6 @@
     "## Execute model"
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "%load_ext autoreload\n",
-    "%autoreload 2"
-   ]
-  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -422,25 +417,7 @@
    "outputs": [],
    "source": [
     "url = 'http://vision.cs.arizona.edu/adarsh/export/demos/data/pi_mtg_demo_unfiltered.pkl'\n",
-    "with urllib.request.urlopen(url) as f:\n",
-    "    sts = pickle.load(f) "
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "slideshow": {
-     "slide_type": "subslide"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "from delphi.subgraphs import *\n",
-    "G = AnalysisGraph.from_statements(sts)\n",
-    "merge_nodes(G, 'food_security', 'food_insecurity', same_polarity=False)\n",
-    "G = get_subgraph_for_concept(G, 'food_insecurity', depth_limit=1)\n",
-    "visualize(G, nodes_to_highlight='food_insecurity')"
+    "sts=pickle.load(get_data_from_url(url))"
    ]
   },
   {
@@ -453,6 +430,7 @@
    },
    "outputs": [],
    "source": [
+    "from delphi.subgraphs import get_subgraph_for_concept_pair\n",
     "G = AnalysisGraph.from_statements(sts)\n",
     "merge_nodes(G, 'food_security', 'food_insecurity', same_polarity=False)\n",
     "G = get_subgraph_for_concept_pair(G, 'drought', 'food_insecurity', cutoff=2)\n",
@@ -798,7 +776,12 @@
    "title_cell": "Table of Contents",
    "title_sidebar": "Contents",
    "toc_cell": false,
-   "toc_position": {},
+   "toc_position": {
+    "height": "calc(100% - 180px)",
+    "left": "10px",
+    "top": "150px",
+    "width": "165px"
+   },
    "toc_section_display": true,
    "toc_window_display": true
   }