def test_summer_length2():
"""
Ensure proper dict of dicts from summer_length
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a.summer_length(swe)
assert type(c[(1, 1)]) == dict
def test_summer_length():
"""
Ensure summer length info is returned in a dictionary
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a.summer_length(swe)
assert type(c) == dict
def test_make_df_type():
"""
Ensure make_df creates a proper dataframe (could be more meaningful)
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
t = a.make_df()
assert type(t) == pd.DataFrame
def test_count_melt():
"""
Ensure dataframe is returned by __count_melt (could be more meaningful?)
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a._count_melt(swe)
assert type(c) == pd.DataFrame
def test_create_splits():
"""
Ensure proper date splits are generated given specific date and size of array
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
years = a.create_year_splits()
assert years == [0, 365, 730, 1095, 1460, 1826, 2191, 2556, 2921, 3000]
def test_melt_date_year():
"""
Ensure a dictionary is returned for counting melt date by year
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
df = a.make_df()
counts = a.melt_date_year(df)
assert type(counts) == dict
def test_mask_year_df():
"""
Ensure proper structure of dataframe given a mask on 1994
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
df = a.make_df()
df = a.mask_year_df(df, 1994)
assert df.iloc[0].time.year == 1994
def test_count_melt_onset_mp():
"""
Note: right now making zero matrix swe values
- should maybe make fixture that has an instantiated analysis object with the swe cube inside
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a.count_melt_onset()
assert type(c) == pd.DataFrame
def test_display_melt_fail2():
"""
Ensure exception is raised when dates are nonsensical
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
melt = a.melt_date_year(a._count_melt(swe))
with pytest.raises(Exception):
a.display_melt_onset_change(melt, 1993, 1944)
def test_create_splits_mid_year():
"""
Make sure that year splits are properly generated when not using first day of year
as start of time series
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 5, 1), swe)
years = a.create_year_splits()
assert years[0] == 121
def test_display_melt():
"""
Ensure proper figure type returned for summer diff map
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
melt = a.melt_date_year(a._count_melt(swe))
fig = a.display_melt_onset_change(melt, 1993, 1995, True)
assert type(fig) == matplotlib.figure.Figure
def test_make_df_time():
"""
Make sure timestamp is correctly created in dataframe
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
t = a.make_df()
print(type(t.time[0]))
assert t.time[0] == pd.Timestamp("1993-01-01 00:00:00")
def test_summer_diff2():
"""
Make sure heatmap is proper size
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a.summer_length(swe)
diff, heatmap = a.summer_diff(c)
assert np.shape(heatmap) == (50, 50)
def test_display_diffmap():
"""
Ensure proper image type returned for summer diff map
"""
swe = np.zeros((3000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a.summer_length(swe)
diff, heatmap = a.summer_diff(c)
im = a.display_summer_change(True)
assert type(im) == matplotlib.image.AxesImage
def test_summer_diff1():
"""
Ensure summer length difference is close to zero on test array
"""
swe = np.zeros((1000, 50, 50))
a = analysis.Analysis(datetime.date(1993, 1, 1), swe)
c = a.summer_length(swe)
diff, heatmap = a.summer_diff(c)
if diff < 0.1:
diff = 0
assert diff == 0
def a():
swe = np.zeros((3000, 50, 50))
return analysis.Analysis(datetime.date(1993, 1, 1), swe)
