def test_battle_create_army():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
# create army
# try normal
with pytest.raises(TypeError):
b.create_army("Hello")
with pytest.raises(TypeError):
b.create_army(pd.DataFrame({"ho": [1, 2], "hi": [2, 3]}))
with pytest.raises(TypeError):
b.create_army(["Clone"])
with pytest.raises(TypeError):
b.create_army(["Clone", 2])
with pytest.raises(ValueError):
b.create_army([("Clone", 2)])
b.create_army([("Droid", 10), ("Clone Trooper", 5)])
with pytest.raises(TypeError):
b.create_army([("Clone Trooper", "hello")])
b.create_army([("B1 battledroid", np.inf)])
# created normally.
b.create_army([("B1 battledroid", 10), ("Clone Trooper", 10)])
return b
def test_apply_position():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
# first, attempt to call without create_army
with pytest.raises(AttributeError):
b.apply_position("gaussian")
# created normally.
b.create_army([("B1 battledroid", 100), ("Clone Trooper", 100)])
# try parameter diffs
with pytest.raises(TypeError):
b.apply_position()
with pytest.raises(TypeError):
b.apply_position(1)
# apply_pos with false distribution
with pytest.raises(ValueError):
b.apply_position("fake")
# normal
b.apply_position("gaussian")
# check distribution - pvalue > 0.05
assert normaltest(b.M_["pos"][:, 0])[1] > 0.05, "P-value for gaussian test > 0.05, fail"
b.apply_position("exp")
# check distribution using KS-test > 0.05
assert kstest(b.M_["pos"][:, 0], "expon")[1] > 0.05, "P-value for KS-test not > 0.05, fail"
b.apply_position("laplace")
assert kstest(b.M_["pos"][:, 1], "laplace")[1] > 0.05, "P-value for KS-test not > 0.05, fail"
#
with pytest.raises(ValueError):
b.apply_position(["fake"])
b.apply_position([("Hello")])
# dist for each unit - works
b.apply_position(["gaussian","uniform"])
assert kstest(b.M_["pos"][:100, 0], "norm")[1] > 0.01, "P-value for KS-test not > 0.05, fail"
assert kstest(b.M_["pos"][-100:, 0], "uniform")[1] > 0.01, "P-value for KS-test not > 0.05, fail"
# we don't test bsm.Distribution objects here
b.apply_position({"name":"gaussian", "loc": 0., "scale": 1.})
# ignore wrong keyword
b.apply_position({"n":"gaussian", "loc":0., "scale":1.})
with pytest.raises(ValueError):
b.apply_position({"name":10, "loc":0., "scale":1.})
with pytest.raises(ValueError):
b.apply_position({"name":str(10), "loc":0., "scale":1.})
with pytest.raises(TypeError):
# wrong loc, scale parameters
b.apply_position({"name":"gaussian", "loc":"hello", "scale":1.})
def test_battle_attributes():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
assert b.M_ is None, "M_ should not be set"
assert type(b.db_) == pd.DataFrame, "db_ must be a pandas.dataframe"
assert b.sim_ is None, "sim_ not set yet"
with pytest.raises(AttributeError):
b.composition_
return b
def test_simulate():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
# cannot simulate before creating an army set
with pytest.raises(AttributeError):
b.simulate()
# define army
b.create_army([("B1 battledroid", 100), ("Clone Trooper", 100)])
assert b.sim_ is None, "no simulation object present"
# no important parameters apart from those passed to simulate_battle
# check return type
F = b.simulate()
# check presense of b.sim_
assert b.sim_ is not None, "simulation object should be present and isnt"
assert type(F) is np.ndarray, "must be of type np.ndarray for F"
def test_simulate():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
# cannot simulate before creating an army set
with pytest.raises(AttributeError):
b.simulate()
# define army
b.create_army([("B1 battledroid", 100), ("Clone Trooper", 100)])
assert b.sim_ is None, "no simulation object present"
# no important parameters apart from those passed to simulate_battle
# check return type
F = b.simulate()
# check presense of b.sim_
assert b.sim_ is not None, "simulation object should be present and isnt"
assert type(F) is pd.DataFrame, "must be of type pd.DataFrame for F"
# check columns
for c in F.columns:
assert c in bsm._simulator_fast.frame_columns(), "column '{}' not in df.columnset".format(c)
def test_set_rolling_ai():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
# set AI before definition
with pytest.raises(AttributeError):
b.set_rolling_ai(["nearest", "random"])
with pytest.raises(AttributeError):
b.set_rolling_ai("nearest")
# created normally.
b.create_army([("B1 battledroid", 100), ("Clone Trooper", 100)])
# call error
with pytest.raises(ValueError):
b.set_rolling_ai("hello")
with pytest.raises(ValueError):
b.set_rolling_ai(["nearest","hello"])
# call normally
b.set_rolling_ai(["nearest", "random"])
# check composition
assert set(b.composition_["rolling_ai"].values) == set(["random","nearest"])
def test_define_battle():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
def battle():
return bsm.Battle("../datasets/starwars-clonewars.csv")
def some_setup():
# battle object requires input file
b = bsm.Battle("../datasets/starwars-clonewars.csv")
# created normally.
b.create_army([("B1 battledroid", 100), ("Clone Trooper", 100)])
return b
