using global shared variable in test (lux-org#144)

.gitignore

@@ -10,5 +10,6 @@ static/
 *.egg-info*
 build/
 .DS_Store
+.idea/
 tests/.coverage
 tests/coverage.xml

tests/conftest.py

@@ -0,0 +1,12 @@
+import pytest
+import pandas as pd
+
+
+@pytest.fixture
+def global_var():
+    pytest.olympics_df = pd.read_csv("https://github.com/lux-org/lux-datasets/blob/master/data/olympic.csv?raw=true")
+    pytest.car_df = pd.read_csv("lux/data/car.csv")
+    url = "https://github.com/lux-org/lux-datasets/blob/master/data/state_timeseries.csv?raw=true"
+    pytest.state_timeseries_df = pd.read_csv(url)
+    pytest.college_df = pd.read_csv("lux/data/college.csv")
+    pytest.census_df = pd.read_csv("https://github.com/lux-org/lux-datasets/blob/master/data/census.csv?raw=true")

tests/test_action.py

@@ -19,16 +19,16 @@
 from lux.vis.Vis import Vis
 
 
-def test_vary_filter_val():
-    df = pd.read_csv("https://github.com/lux-org/lux-datasets/blob/master/data/olympic.csv?raw=true")
+def test_vary_filter_val(global_var):
+    df = pytest.olympics_df
     vis = Vis(["Height", "SportType=Ball"], df)
     df.set_intent_as_vis(vis)
     df._repr_html_()
     assert len(df.recommendation["Filter"]) == len(df["SportType"].unique()) - 1
 
 
-def test_filter_inequality():
-    df = pd.read_csv("lux/data/car.csv")
+def test_filter_inequality(global_var):
+    df = pytest.car_df
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
 
     df.set_intent(
@@ -48,9 +48,9 @@ def test_filter_inequality():
     assert fltr_clause.value == 10
 
 
-def test_generalize_action():
+def test_generalize_action(global_var):
     # test that generalize action creates all unique visualizations
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     df["Year"] = pd.to_datetime(
         df["Year"], format="%Y"
     )  # change pandas dtype for the column "Year" to datetype
@@ -72,9 +72,8 @@ def test_generalize_action():
     assert check1 and check2 and check3
 
 
-def test_row_column_group():
-    url = "https://github.com/lux-org/lux-datasets/blob/master/data/state_timeseries.csv?raw=true"
-    df = pd.read_csv(url)
+def test_row_column_group(global_var):
+    df = pytest.state_timeseries_df
     df["Date"] = pd.to_datetime(df["Date"])
     tseries = df.pivot(index="State", columns="Date", values="Value")
     # Interpolating missing values
@@ -85,8 +84,8 @@ def test_row_column_group():
     assert list(tseries.recommendation.keys()) == ["Row Groups", "Column Groups"]
 
 
-def test_groupby():
-    df = pd.read_csv("lux/data/college.csv")
+def test_groupby(global_var):
+    df = pytest.college_df
     groupbyResult = df.groupby("Region").sum()
     groupbyResult._repr_html_()
     assert list(groupbyResult.recommendation.keys()) == ["Column Groups"]
@@ -159,17 +158,17 @@ def test_crosstab():
     assert list(result.recommendation.keys()) == ["Row Groups", "Column Groups"]
 
 
-def test_custom_aggregation():
+def test_custom_aggregation(global_var):
     import numpy as np
 
-    df = pd.read_csv("lux/data/college.csv")
+    df = pytest.college_df
     df.set_intent(["HighestDegree", lux.Clause("AverageCost", aggregation=np.ptp)])
     df._repr_html_()
     assert list(df.recommendation.keys()) == ["Enhance", "Filter", "Generalize"]
 
 
-def test_year_filter_value():
-    df = pd.read_csv("lux/data/car.csv")
+def test_year_filter_value(global_var):
+    df = pytest.car_df
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     df.set_intent(["Acceleration", "Horsepower"])
     df._repr_html_()

tests/test_compiler.py

@@ -19,9 +19,9 @@
 from lux.vis.VisList import VisList
 
 
-def test_underspecified_no_vis(test_recs):
+def test_underspecified_no_vis(global_var, test_recs):
     no_vis_actions = ["Correlation", "Distribution", "Occurrence", "Temporal"]
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     test_recs(df, no_vis_actions)
     assert len(df.current_vis) == 0
 
@@ -31,9 +31,9 @@ def test_underspecified_no_vis(test_recs):
     assert len(df.current_vis) == 0
 
 
-def test_underspecified_single_vis(test_recs):
+def test_underspecified_single_vis(global_var, test_recs):
     one_vis_actions = ["Enhance", "Filter", "Generalize"]
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     df.set_intent([lux.Clause(attribute="MilesPerGal"), lux.Clause(attribute="Weight")])
     test_recs(df, one_vis_actions)
     assert len(df.current_vis) == 1
@@ -75,8 +75,8 @@ def test_underspecified_single_vis(test_recs):
 # 	df.set_intent([lux.Clause(attribute ="?", data_model="measure"), lux.Clause(attribute ="?", data_model="measure")])
 # 	assert len(df.current_vis) == len([vis.get_attr_by_data_model("measure") for vis in df.current_vis]) #should be 25
 # 	test_recs(df, multiple_vis_actions)
-def test_set_intent_as_vis(test_recs):
-    df = pd.read_csv("lux/data/car.csv")
+def test_set_intent_as_vis(global_var, test_recs):
+    df = pytest.car_df
     df._repr_html_()
     vis = df.recommendation["Correlation"][0]
     df.intent = vis
@@ -95,19 +95,19 @@ def test_recs_function(df, actions):
     return test_recs_function
 
 
-def test_parse():
-    df = pd.read_csv("lux/data/car.csv")
+def test_parse(global_var):
+    df = pytest.car_df
     vlst = VisList([lux.Clause("Origin=?"), lux.Clause(attribute="MilesPerGal")], df)
     assert len(vlst) == 3
 
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     vlst = VisList([lux.Clause("Origin=?"), lux.Clause("MilesPerGal")], df)
     assert len(vlst) == 3
 
 
-def test_underspecified_vis_collection_zval():
+def test_underspecified_vis_collection_zval(global_var):
     # check if the number of charts is correct
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     vlst = VisList(
         [
             lux.Clause(attribute="Origin", filter_op="=", value="?"),
@@ -123,11 +123,11 @@ def test_underspecified_vis_collection_zval():
     # assert len(vlst) == 8
 
 
-def test_sort_bar():
+def test_sort_bar(global_var):
     from lux.processor.Compiler import Compiler
     from lux.vis.Vis import Vis
 
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     vis = Vis(
         [
             lux.Clause(attribute="Acceleration", data_model="measure", data_type="quantitative"),
@@ -138,7 +138,7 @@ def test_sort_bar():
     assert vis.mark == "bar"
     assert vis._inferred_intent[1].sort == ""
 
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     vis = Vis(
         [
             lux.Clause(attribute="Acceleration", data_model="measure", data_type="quantitative"),
@@ -150,8 +150,8 @@ def test_sort_bar():
     assert vis._inferred_intent[1].sort == "ascending"
 
 
-def test_specified_vis_collection():
-    df = pd.read_csv("lux/data/car.csv")
+def test_specified_vis_collection(global_var):
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
 
@@ -181,8 +181,8 @@ def test_specified_vis_collection():
     assert "Origin = Europe" not in chart_titles
 
 
-def test_specified_channel_enforced_vis_collection():
-    df = pd.read_csv("lux/data/car.csv")
+def test_specified_channel_enforced_vis_collection(global_var):
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     visList = VisList(
@@ -193,9 +193,9 @@ def test_specified_channel_enforced_vis_collection():
         check_attribute_on_channel(vis, "MilesPerGal", "x")
 
 
-def test_autoencoding_scatter():
+def test_autoencoding_scatter(global_var):
     # No channel specified
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     vis = Vis([lux.Clause(attribute="MilesPerGal"), lux.Clause(attribute="Weight")], df)
@@ -234,9 +234,9 @@ def test_autoencoding_scatter():
         )
 
 
-def test_autoencoding_histogram():
+def test_autoencoding_histogram(global_var):
     # No channel specified
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     vis = Vis([lux.Clause(attribute="MilesPerGal", channel="y")], df)
@@ -247,8 +247,8 @@ def test_autoencoding_histogram():
     assert vis.get_attr_by_channel("y")[0].attribute == "Record"
 
 
-def test_autoencoding_line_chart():
-    df = pd.read_csv("lux/data/car.csv")
+def test_autoencoding_line_chart(global_var):
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     vis = Vis([lux.Clause(attribute="Year"), lux.Clause(attribute="Acceleration")], df)
@@ -287,8 +287,8 @@ def test_autoencoding_line_chart():
         )
 
 
-def test_autoencoding_color_line_chart():
-    df = pd.read_csv("lux/data/car.csv")
+def test_autoencoding_color_line_chart(global_var):
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     intent = [
@@ -302,8 +302,8 @@ def test_autoencoding_color_line_chart():
     check_attribute_on_channel(vis, "Origin", "color")
 
 
-def test_autoencoding_color_scatter_chart():
-    df = pd.read_csv("lux/data/car.csv")
+def test_autoencoding_color_scatter_chart(global_var):
+    df = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     vis = Vis(
@@ -327,10 +327,10 @@ def test_autoencoding_color_scatter_chart():
     check_attribute_on_channel(vis, "Acceleration", "color")
 
 
-def test_populate_options():
+def test_populate_options(global_var):
     from lux.processor.Compiler import Compiler
 
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     df.set_intent([lux.Clause(attribute="?"), lux.Clause(attribute="MilesPerGal")])
     col_set = set()
     for specOptions in Compiler.populate_wildcard_options(df._intent, df)["attributes"]:
@@ -355,8 +355,8 @@ def test_populate_options():
     )
 
 
-def test_remove_all_invalid():
-    df = pd.read_csv("lux/data/car.csv")
+def test_remove_all_invalid(global_var):
+    df = pytest.car_df
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
     # with pytest.warns(UserWarning,match="duplicate attribute specified in the intent"):
     df.set_intent(

tests/test_config.py

@@ -48,8 +48,8 @@ def contain_horsepower(df):
     return df
 
 
-def test_default_actions_registered():
-    df = pd.read_csv("lux/data/car.csv")
+def test_default_actions_registered(global_var):
+    df = pytest.car_df
     df._repr_html_()
     assert "Distribution" in df.recommendation
     assert len(df.recommendation["Distribution"]) > 0
@@ -91,13 +91,13 @@ def test_no_validator():
     assert "bars" in df.recommendation
 
 
-def test_invalid_function():
+def test_invalid_function(global_var):
     df = pd.read_csv("lux/data/car.csv")
     with pytest.raises(ValueError, match="Value must be a callable"):
         lux.register_action("bars", "not a Callable")
 
 
-def test_invalid_validator():
+def test_invalid_validator(global_var):
     df = pd.read_csv("lux/data/car.csv")
 
     def random_categorical(ldf):
@@ -136,14 +136,14 @@ def test_remove_action():
     )
 
 
-def test_remove_invalid_action():
-    df = pd.read_csv("lux/data/car.csv")
+def test_remove_invalid_action(global_var):
+    df = pytest.car_df
     with pytest.raises(ValueError, match="Option 'bars' has not been registered"):
         lux.remove_action("bars")
 
 
-def test_remove_default_actions():
-    df = pd.read_csv("lux/data/car.csv")
+def test_remove_default_actions(global_var):
+    df = pytest.car_df
     df._repr_html_()
 
     lux.remove_action("Distribution")
@@ -178,8 +178,8 @@ def test_remove_default_actions():
 
 
 # TODO: This test does not pass in pytest but is working in Jupyter notebook.
-# def test_plot_setting():
-# 	df = pd.read_csv("lux/data/car.csv")
+# def test_plot_setting(global_var):
+# 	df = pytest.car_df
 # 	df["Year"] = pd.to_datetime(df["Year"], format='%Y')
 # 	def change_color_add_title(chart):
 # 		chart = chart.configure_mark(color="green") # change mark color to green

tests/test_dates.py

@@ -20,8 +20,8 @@
 from lux.executor.PandasExecutor import PandasExecutor
 
 
-def test_dateformatter():
-    ldf = pd.read_csv("lux/data/car.csv")
+def test_dateformatter(global_var):
+    ldf = pytest.car_df
     # change pandas dtype for the column "Year" to datetype
     ldf["Year"] = pd.to_datetime(ldf["Year"], format="%Y")
     timestamp = np.datetime64("2019-08-26")
@@ -37,8 +37,8 @@ def test_dateformatter():
     assert date_utils.date_formatter(timestamp, ldf) == "2019-8-26"
 
 
-def test_period_selection():
-    ldf = pd.read_csv("lux/data/car.csv")
+def test_period_selection(global_var):
+    ldf = pytest.car_df
     ldf["Year"] = pd.to_datetime(ldf["Year"], format="%Y")
 
     ldf["Year"] = pd.DatetimeIndex(ldf["Year"]).to_period(freq="A")
@@ -56,8 +56,8 @@ def test_period_selection():
     assert all(ldf.current_vis[2].data.columns == ["Year", "Acceleration"])
 
 
-def test_period_filter():
-    ldf = pd.read_csv("lux/data/car.csv")
+def test_period_filter(global_var):
+    ldf = pytest.car_df
     ldf["Year"] = pd.to_datetime(ldf["Year"], format="%Y")
 
     ldf["Year"] = pd.DatetimeIndex(ldf["Year"]).to_period(freq="A")
@@ -70,9 +70,9 @@ def test_period_filter():
     assert isinstance(ldf.recommendation["Filter"][2]._inferred_intent[2].value, pd.Period)
 
 
-def test_period_to_altair():
+def test_period_to_altair(global_var):
     chart = None
-    df = pd.read_csv("lux/data/car.csv")
+    df = pytest.car_df
     df["Year"] = pd.to_datetime(df["Year"], format="%Y")
 
     df["Year"] = pd.DatetimeIndex(df["Year"]).to_period(freq="A")