Add Vega-Lite Auto Examples

altair/examples/__init__.py

@@ -1,15 +1,19 @@
 """Code for Example Plots"""
+import os
+import json
 
+JSON_DIR = os.path.abspath(os.path.join(os.path.dirname(__file__), 'json'))
 
-def load_example(name):
-    import os
-    import json
 
-    JSON_DIR = os.path.join(os.path.dirname(__file__), 'json')
-    if name not in os.listdir(JSON_DIR):
-        raise ValueError("Example='{0}' not valid.".format(name))
+def load_example(filename):
+    """Load the JSON string corresponding to the given filename"""
+    if filename not in os.listdir(JSON_DIR):
+        raise ValueError("Example='{0}' not valid.".format(filename))
 
-    with open(os.path.join(JSON_DIR, name)) as f:
-        schema = json.load(f)
+    with open(os.path.join(JSON_DIR, filename)) as f:
+        return json.load(f)
 
-    return schema
+
+def iter_examples():
+    for filename in os.listdir(JSON_DIR):
+        yield filename, load_example(filename)

altair/examples/tests/test_examples.py

@@ -4,56 +4,19 @@
 
 import pytest
 
+from .. import iter_examples
 from ...datasets import connection_ok, URLError
 from ... import examples
 from ... import *
 
 
-def iter_submodules(package, include_packages=False):
-    """Iterate importable submodules of a module"""
-    prefix = package.__name__ + "."
-    for importer, modname, ispkg in pkgutil.iter_modules(package.__path__,
-                                                         prefix):
-        if not include_packages and ispkg:
-            continue
-        yield modname
-
-
-def iter_json_examples():
-    json_dir = os.path.abspath(os.path.join(os.path.dirname(__file__),
-                                            '..', 'json'))
-    for json_file in os.listdir(json_dir):
-        yield os.path.join(json_dir, json_file)
-
-
-@pytest.mark.skipif(not connection_ok(), reason='No Internet Connection')
-@pytest.mark.parametrize('modname', iter_submodules(examples))
-def test_examples_output(modname):
-    # Don't compare data here, as the JSON representation is tricky
-    example = __import__(modname, fromlist="dummy")
-    v = example.v
-    vdict = v.to_dict(data=False)
-
-    # test that output matches expected output
-    assert vdict == example.expected_output
-
-    # test from_dict methods
-    v2 = Layer.from_dict(example.expected_output)
-    assert v2.to_dict() == vdict
-
-    # test to_altair methods
-    v3 = eval(v.to_altair())
-    assert v3.to_dict() == vdict
-
-
 @pytest.mark.skipif(not connection_ok(), reason='No Internet Connection')
-@pytest.mark.parametrize('json_path', iter_json_examples())
-def test_json_examples_round_trip(json_path):
+@pytest.mark.parametrize('example', iter_examples())
+def test_json_examples_round_trip(example):
     """
     Test that Altair correctly round-trips JSON with to_dict() and to_altair()
     """
-    with open(json_path) as f:
-        json_dict = json.load(f)
+    filename, json_dict = example
 
     v = Layer.from_dict(json_dict)
     v_dict = v.to_dict()

altair/tests/test_api.py

@@ -8,26 +8,6 @@
 from .. import *
 from ..utils import parse_shorthand, infer_vegalite_type
 
-EXAMPLE_JSON = """
-{
-  "data": {"url": "http://vega.github.io/vega-lite/data/unemployment-across-industries.json"},
-  "mark": "area",
-  "encoding": {
-    "x": {
-      "timeUnit": "yearmonth", "field": "date", "type": "temporal",
-      "scale": {"nice": "month"},
-      "axis": {"axisWidth": 0, "format": "%Y", "labelAngle": 0, "tickSize": 0}
-    },
-    "y": {
-      "aggregate": "sum", "field": "count","type": "quantitative",
-      "axis": false
-    },
-    "color": {"field":"series", "type":"nominal", "scale":{"range": "category20b"}}
-  },
-  "config": {"cell": {"width": 300, "height": 200}, "mark": {"stacked": "center"}}
-}
-"""
-
 
 def test_encode_update():
     # Test that encode updates rather than overwrites
@@ -92,10 +72,3 @@ def test_to_altair_stocks():
     layer2 = eval(code)
 
     assert layer.to_dict() == layer2.to_dict()
-
-
-def test_from_json():
-    layer = Layer.from_json(EXAMPLE_JSON)
-    D1 = json.loads(EXAMPLE_JSON)
-    D2 = layer.to_dict()
-    assert D1 == D2

altair/utils.py

@@ -171,6 +171,7 @@ def sanitize_dataframe(df):
     * Raise ValueError is it has a hierarchical index.
     * Convert categoricals to strings.
     * Convert np.int dtypes to Python int objects
+    * Convert DateTime dtypes into appropriate string representations
     """
     import pandas as pd
     import numpy as np
@@ -188,6 +189,10 @@ def sanitize_dataframe(df):
             df[col_name] = df[col_name].astype(str)
         elif np.issubdtype(dtype, np.integer):
             df[col_name] = np.array(list(map(int, df[col_name])), dtype=object)
+        elif str(dtype) == 'datetime64[ns]':
+            # this is the format used in Vega-Lite example area.json
+            to_str = lambda x: x.strftime('%Y-%m-%dT%H:%M:%S.%f')[:-3] + 'Z'
+            df[col_name] = df[col_name].apply(to_str)
     return df
 
 def dataframe_to_json(df):

notebooks/Index.ipynb

@@ -270,7 +270,25 @@
     "3. [Line Plots](LinePlots.ipynb)\n",
     "4. [Area Plots](AreaPlots.ipynb)\n",
     "5. [Exploring Cars Data](Cars.ipynb)\n",
-    "6. [PairGrid](PairGrid.ipynb)\n",
+    "6. [PairGrid](PairGrid.ipynb)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Examples from Vega-Lite\n",
+    "\n",
+    "The Vega-Lite package includes dozens of example plots; examples of how to create them in Alair are shown here:\n",
+    "\n",
+    "1. [Vega-Lite Example Plots](auto_examples/Index.ipynb)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## More Documentation\n",
     "\n",
     "The remainder of the documentation is in a set of notebooks diving into the use of Altair:\n",
     "\n",

notebooks/auto_examples/Index.ipynb

@@ -0,0 +1,92 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Auto-Generated Altair Examples\n",
+    "\n",
+    "All the following notebooks are auto-generated from the example specifications\n",
+    "in the [Vega-Lite](http://vega.github.io/vega-lite/) project.\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "- [Area](area.ipynb): *--*\n",
+    "- [Area Vertical](area_vertical.ipynb): *Area chart showing weight of cars over time (vertical).*\n",
+    "- [Bar 1D](bar_1d.ipynb): *--*\n",
+    "- [Bar Aggregate](bar_aggregate.ipynb): *A bar chart showing the US population distribution of age groups in 2000.*\n",
+    "- [Bar Aggregate Size](bar_aggregate_size.ipynb): *A bar chart showing the US population distribution of age groups in 2000.*\n",
+    "- [Bar Aggregate Vertical](bar_aggregate_vertical.ipynb): *--*\n",
+    "- [Bar Grouped](bar_grouped.ipynb): *--*\n",
+    "- [Bar Grouped Horizontal](bar_grouped_horizontal.ipynb): *--*\n",
+    "- [Bar Layered Transparent](bar_layered_transparent.ipynb): *A bar chart showing the US population distribution of age groups and gender in 2000.*\n",
+    "- [Bar Yearmonth](bar_yearmonth.ipynb): *Temperature in Seattle as a bar chart with yearmonth time unit.*\n",
+    "- [Circle](circle.ipynb): *--*\n",
+    "- [Histogram](histogram.ipynb): *--*\n",
+    "- [Line](line.ipynb): *Google's stock price over time.*\n",
+    "- [Line Color](line_color.ipynb): *Stock prices of 5 Tech Companies Over Time.*\n",
+    "- [Line Detail](line_detail.ipynb): *Stock prices of 5 Tech Companies Over Time.*\n",
+    "- [Line Monotone](line_monotone.ipynb): *--*\n",
+    "- [Line Month](line_month.ipynb): *--*\n",
+    "- [Line Slope](line_slope.ipynb): *Slope graph showing the change in yield for different barley sites. It shows the error in the year labels for the Morris site.*\n",
+    "- [Line Step](line_step.ipynb): *Google's stock price over time.*\n",
+    "- [Point 1D](point_1d.ipynb): *--*\n",
+    "- [Point Dot Timeunit Color](point_dot_timeunit_color.ipynb): *--*\n",
+    "- [Point Filled](point_filled.ipynb): *--*\n",
+    "- [Scatter](scatter.ipynb): *A scatterplot showing horsepower and miles per gallons for various cars.*\n",
+    "- [Scatter Aggregate Detail](scatter_aggregate_detail.ipynb): *A scatterplot showing average horsepower and displacement for cars from different origins.*\n",
+    "- [Scatter Binned](scatter_binned.ipynb): *--*\n",
+    "- [Scatter Binned Color](scatter_binned_color.ipynb): *A scatterplot showing horsepower and miles per gallons with binned acceleration on color.*\n",
+    "- [Scatter Binned Size](scatter_binned_size.ipynb): *A scatterplot showing horsepower and miles per gallons with binned acceleration on size.*\n",
+    "- [Scatter Bubble](scatter_bubble.ipynb): *A bubbleplot showing horsepower on x, miles per gallons on y, and binned acceleration on size.*\n",
+    "- [Scatter Color](scatter_color.ipynb): *--*\n",
+    "- [Scatter Color Custom](scatter_color_custom.ipynb): *--*\n",
+    "- [Scatter Color Order](scatter_color_order.ipynb): *--*\n",
+    "- [Scatter Color Ordinal](scatter_color_ordinal.ipynb): *--*\n",
+    "- [Scatter Color Ordinal Custom](scatter_color_ordinal_custom.ipynb): *--*\n",
+    "- [Scatter Color Quantitative](scatter_color_quantitative.ipynb): *--*\n",
+    "- [Scatter Color Shape Constant](scatter_color_shape_constant.ipynb): *--*\n",
+    "- [Scatter Colored With Shape](scatter_colored_with_shape.ipynb): *A scatterplot showing horsepower and miles per gallons.*\n",
+    "- [Scatter Connected](scatter_connected.ipynb): *--*\n",
+    "- [Square](square.ipynb): *--*\n",
+    "- [Stacked Area](stacked_area.ipynb): *Area chart showing weight of cars over time.*\n",
+    "- [Stacked Area Normalize](stacked_area_normalize.ipynb): *--*\n",
+    "- [Stacked Area Ordinal](stacked_area_ordinal.ipynb): *--*\n",
+    "- [Stacked Area Stream](stacked_area_stream.ipynb): *--*\n",
+    "- [Stacked Bar 1D](stacked_bar_1d.ipynb): *--*\n",
+    "- [Stacked Bar H](stacked_bar_h.ipynb): *--*\n",
+    "- [Stacked Bar H Order](stacked_bar_h_order.ipynb): *--*\n",
+    "- [Stacked Bar Normalize](stacked_bar_normalize.ipynb): *--*\n",
+    "- [Stacked Bar Population](stacked_bar_population.ipynb): *A bar chart showing the US population distribution of age groups and gender in 2000.*\n",
+    "- [Stacked Bar V](stacked_bar_v.ipynb): *--*\n",
+    "- [Stacked Bar Weather](stacked_bar_weather.ipynb): *--*\n",
+    "- [Text Scatter Colored](text_scatter_colored.ipynb): *--*\n",
+    "- [Text Table Heatmap](text_table_heatmap.ipynb): *--*\n",
+    "- [Tick Dot](tick_dot.ipynb): *--*\n",
+    "- [Tick Dot Thickness](tick_dot_thickness.ipynb): *--*\n",
+    "- [Tick Strip](tick_strip.ipynb): *Shows the relationship between horsepower and the numbver of cylinders using tick marks.*\n",
+    "- [Trellis Bar](trellis_bar.ipynb): *A trellis bar chart showing the US population distribution of age groups and gender in 2000.*\n",
+    "- [Trellis Bar Histogram](trellis_bar_histogram.ipynb): *--*\n",
+    "- [Trellis Barley](trellis_barley.ipynb): *The Trellis display by Becker et al. helped establish small multiples as a “powerful mechanism for understanding interactions in studies of how a response depends on explanatory variables”. Here we reproduce a trellis of Barley yields from the 1930s, complete with main-effects ordering to facilitate comparison.*\n",
+    "- [Trellis Row Column](trellis_row_column.ipynb): *--*\n",
+    "- [Trellis Scatter](trellis_scatter.ipynb): *--*\n",
+    "- [Trellis Scatter Binned Row](trellis_scatter_binned_row.ipynb): *A trellis scatterplot showing Horsepower and Miles per gallons, faceted by binned values of Acceleration.*\n",
+    "- [Trellis Stacked Bar](trellis_stacked_bar.ipynb): *--*"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "env": {},
+   "language": "",
+   "name": "python3"
+  },
+  "language": "python"
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}