def test_multiply():
df1 = load_series(generator="linear")
df2 = load_series(generator="harmonic")
df = process(function="multiply",
df_first=df1,
df_second=df2,
first_col="u",
second_col="u")
assert df.columns == ["u"]
def test_data_create():
component_first_df = load_series(generator="harmonic", points=10000)
component_second_df = load_series(generator="white_noise", points=10000)
signal_df = component_first_df["u"] + component_second_df["u"]
# saving
file_path_3 = (
os.path.dirname(os.path.dirname(os.path.abspath(__file__))) +
"/pure_data/gauss_additiv.txt")
signal_df.to_csv(file_path_3, index=False, header=None)
assert not signal_df.empty
def test_valid_dfa_handle():
df = load_series(path=abs_path + "/pure_data/close_mod.txt")
df = process(function="profile", df=df)
df = process(function="dfa_extended", df=df)
dfa_handler(df)
assert os.listdir(csv_dir)
assert len(os.listdir(img_dir)) == 5
def test_dfa_many():
df = load_series(path=file_path)
df = process(function="profile", df=df)
df = process(function="dfa_extended", df=df)
assert not df["dfa_lags"].empty
assert not df["dfa_transform"].empty
assert not df["dfa_ext_transform"].empty
def test_diff_solution_seq_handle_correctly():
diff_solution_df = load_series(
generator="diff_sol",
t=[1, 1, 1, 0.1, 0.1, 0.1],
f=[f_x, f_y, f_z, f_u, f_v, f_w],
)
diff_solution_df.columns = ["t", "x", "y", "z", "u", "v", "w"]
# saving
file_path_2 = (
os.path.dirname(os.path.dirname(os.path.abspath(__file__))) +
"/pure_data/diff_sol.txt")
diff_solution_df["u"].to_csv(file_path_2, index=False, header=None)
assert not diff_solution_df.empty
def build_dfa_graphics(series_dir, path, scale) -> None:
"""
web-interface supporting method
it contains all the step for DFA-transform building
and saves processing result
saves graphics of profile, time series
alpha and betta coefficient dependence
"""
del_old_files("/static/images")
del_old_files("/dataframes")
# dfa building
df = load_series(path=base_dir + "/" + series_dir + "/" + path)
df = process(function="profile", df=df)
df = process(function="dfa_extended", df=df)
# routing files
csv_path = base_dir + "/dataframes/dataframe.csv"
orig_img_path = base_dir + "/static/images/orig.png"
profile_img_path = base_dir + "/static/images/profile.png"
dfa_img_path = base_dir + "/static/images/dfa.png"
dfa_ext_img_path = base_dir + "/static/images/dfa_ext.png"
dfa_many_img_path = base_dir + "/static/images/dfa_many.png"
# saving results
df.to_csv(csv_path, index=False)
sns.lineplot(data=df["u"][scale[0]:scale[1]])\
.get_figure().savefig(orig_img_path)
plt.clf()
sns.lineplot(data=df["profile"]).get_figure().savefig(profile_img_path)
plt.clf()
sns.lineplot(x=df["dfa_lags"], y=df["dfa_ext_transform"]).get_figure().savefig(
dfa_ext_img_path
)
plt.clf()
sns.lineplot(x=df["dfa_lags"], y=df["dfa_transform"]).get_figure().savefig(
dfa_img_path
)
sns.lineplot(x=df["dfa_lags"], y=df["dfa_ext_transform"]).get_figure().savefig(
dfa_many_img_path
)
plt.clf()
def build_cd_dfa_graphics(series_dir, path, scale):
"""
new algorithm - cd dfa build
"""
del_old_files("/static/images")
del_old_files("/dataframes")
# time series and cD coefficient obtaining
df = load_series(path=base_dir + "/" + series_dir + "/" + path)
order = "db8"
cA, cD = pywt.dwt(df, order)
cd_df = pd.DataFrame({"u": cD.transpose()[-1]})
# dfa building
df = process(function="profile", df=df)
df = process(function="dfa_extended", df=df)
cd_df = process(function="profile", df=cd_df)
cd_df = process(function="dfa_extended", df=cd_df)
# routing files
orig_img_path = base_dir + "/static/images/time_series.png"
cd_img_path = base_dir + "/static/images/cD.png"
cd_dfa_img_path = base_dir + "/static/images/DFA_cD.png"
ts_path = base_dir + "/dataframes/time_series.csv"
cd_path = base_dir + "/dataframes/cD.csv"
# saving results
df.to_csv(ts_path, index=False)
cd_df.to_csv(cd_path, index=False)
sns.lineplot(data=df["u"][scale[0]:scale[1]])\
.get_figure().savefig(orig_img_path)
plt.clf()
sns.lineplot(data=cd_df["u"]).get_figure().savefig(cd_img_path)
plt.clf()
sns.lineplot(x=cd_df["dfa_lags"], y=cd_df["dfa_transform"]).get_figure().savefig(
cd_dfa_img_path
)
plt.clf()
def test_existing_raw_file_aggregation():
assert not load_series(path=file_path).empty
def save_ts(upload_dir, file_name): # pre-index()
process(function="filtering",
df=load_series(path="pure_data/linear.txt"))\
.to_csv(f"{upload_dir}/{file_name}", header=None, index=False)
def test_dfa_points():
df = load_series(path=file_path)
df = process(function="profile", df=df)
df = process(function="dfa_extended", df=df)
assert (len(df["dfa_lags"]) == len(df["dfa_transform"]) == len(
df["dfa_ext_transform"]))
def test_akf():
lags_points = 10
df = load_series(generator="harmonic")
df = process(function="akf", df=df, lags=lags_points)
assert not df["akf"].empty
def test_approximation():
df = load_series(path=file_path)
df = process(function="approx", df=df)
assert not df["approx"].empty
def test_profile():
df = load_series(path=file_path)
df = process(function="profile", df=df)
assert not df["profile"].empty
def test_integrate():
df = load_series(path=file_path)
df = process(function="integrate", df=df)
assert not df["integrate"].empty
