def func_resample(args):
output_path = args.outputPath
input_path = args.inputPath
resample_time = "W"
print("Resample input data to:")
print("Input path : {}".format(input_path))
print("Output path : {}".format(output_path))
print("Resample to : {}".format(resample_time))
files = [input_path
] if os.path.isfile(input_path) else os.listdir(input_path)
with tqdm(total=len(files)) as pbar:
for file in files:
full_path = os.path.join(input_path, file)
df = Loader.get_historical_data(full_path,
Source.LocalData,
time_frame=TimeFrame.Daily)
df = df.set_index("date")
result_df = resample_data(df, resample_time)
full_path = os.path.join(output_path, file)
result_df.to_csv(full_path, index=True, header=True)
pbar.set_description("done %s" % file)
pbar.update(1)
def test_that_local_data_source_return_right_schema_dataframe():
base_dir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
test_file_path = os.path.join(base_dir, "sample_data.csv")
df = Loader.get_historical_data(
test_file_path, Source.LocalData, base_date, TimeFrame.Weekly
)
assert (df.columns.sort_values() == data_source_required_columns).all()
def get_hyperwave_from_source(source, symbol, ticker):
pbar.set_description("Processing %s" % symbol)
df_raw_data = Loader.get_historical_data(
symbol, source, time_frame=TimeFrame.Weekly)
(df_hull_hyperwave, hw_phases_temp,
hyperwave) = hw.get_hyperwave(df_raw_data)
hyperwave["ticker"] = ticker
pbar.update(1)
return hyperwave
def get_hyperwave_from_source(source, symbol, ticker):
pbar.set_description("Processing %s" % symbol)
df_raw_data = Loader.get_historical_data(
symbol, source, time_frame=TimeFrame.Weekly)
result = (consensio.get_consonsio(df_raw_data)[[
"date", "consensio"
]].rename(columns={
"consensio": ticker
}).reset_index(drop=True).set_index("date"))
pbar.update(1)
return result
def task_fetch_symbol(source: Source, symbol: str, output_path: str,
timeframe: TimeFrame):
try:
data = Loader.get_historical_data(symbol, source, time_frame=timeframe)
data.to_csv(
output_path,
columns=["date", "close", "high", "low", "open", "volume"],
index=False,
header=True,
)
except Exception as e:
return (e, symbol)
return (None, symbol)
def test_that_investopedia_source_return_right_schema_dataframe(monkeypatch):
def mock_pandas_read_html(url_symbol, header, parse_dates):
data = [
{
"Date": "Oct 01, 2018",
"Open": 227.95,
"High": 229.42,
"Low": 226.35,
"Adj. Close": 227.26,
"Volume": 23600802,
}
]
return [pd.DataFrame(data)]
monkeypatch.setattr(pd, "read_html", mock_pandas_read_html)
df = Loader.get_historical_data("AAPL", Source.Investopedia, base_date)
assert (df.columns.sort_values() == data_source_required_columns).all()
def test_consensio(source_path: str, result_file_path: str, comment: str):
root_test, result_file_name = os.path.split(result_file_path)
df_source = Loader.get_historical_data(source_path,
Source.LocalData,
time_frame=TimeFrame.Weekly)
df_consensio = get_consonsio(df_source,
"close",
low_ma=3,
mid_ma=7,
high_ma=30)
if not os.path.exists(os.path.join(root_test, "results")):
os.mkdir(os.path.join(root_test, "results"))
result_output_path = os.path.join(root_test, "results", result_file_name)
df_consensio.sort_values("date", ascending=True).to_csv(result_output_path,
header=True,
index=False)
if not os.path.exists(result_file_path):
print("Result file doesnt exist")
return
df_result = pd.read_csv(result_file_path,
header=0,
parse_dates=["date"],
infer_datetime_format=True).set_index("date")
df_result["date"] = df_result.index
df_result = (df_result.reset_index(drop=True).sort_values(
"date", ascending=True).reset_index())
df_consensio = (df_consensio.reset_index(drop=True).sort_values(
"date", ascending=True).reset_index())
pd.testing.assert_frame_equal(
df_result.sort_values("date", ascending=True)[["date", "consensio"]],
df_consensio.sort_values("date", ascending=True)[["date",
"consensio"]],
check_less_precise=True,
check_like=True,
)
def func_check(args):
hw = Hyperwave.get_standard_hyperwave()
df_raw_data = Loader.get_historical_data(
args.symbol, args.source, time_frame=TimeFrame.Weekly
)
(df_hull_hyperwave, hw_phases_temp, hyperwave) = hw.get_hyperwave(df_raw_data)
if args.outputType == "display":
print(df_hull_hyperwave)
print(hw_phases_temp)
print(hyperwave)
elif args.outputType == "file":
output_path = args.outputPath
if not os.path.exists(output_path):
print(
"The path {} doesn't exist. Please select a valid outputPath".format(
output_path
)
)
file_name = "{}_{}_{:%Y%m%d}".format(
args.source, args.symbol, datetime.datetime.now()
)
raw_data_file_name = "{}.csv".format(file_name)
hull_file_name = "{}.hull.csv".format(file_name)
hw_file_name = "{}.hw.csv".format(file_name)
df_raw_data.to_csv(
os.path.join(output_path, raw_data_file_name),
columns=["date", "close", "high", "low", "open"],
index=False,
header=True,
)
df_hull_hyperwave.to_csv(
os.path.join(output_path, hull_file_name), header=True, index=False
)
hyperwave.to_csv(
os.path.join(output_path, hw_file_name), header=True, index=False
)
def test_that_cryptocompare_source_return_right_schema_dataframeHyperwave_Path_Finder():
df = Loader.get_historical_data(
"BTC-USD", Source.CryptoCompare, base_date, TimeFrame.Weekly
)
assert (df.columns.sort_values() == data_source_required_columns).all()
def test_that_source_available_is_equal_to_5():
sources = Loader.get_available_sources()
assert len(sources) == 5
def test_tiingo_can_load_data():
df = Loader.get_historical_data("AAPL", Source.Tiingo, base_date)
assert (df.columns.sort_values() == data_source_required_columns).all()
def test_stooq_fail_load_apple():
with pytest.raises(NameError):
df = Loader.get_historical_data("AAPL", Source.Stooq, base_date)
def test_that_we_can_load_from_stooq():
df = Loader.get_historical_data("BTCUSD", Source.Stooq, base_date)
assert (df.columns.sort_values() == data_source_required_columns).all()
def test_that_local_data_source_raise_filenotfound_error():
with pytest.raises(FileNotFoundError):
df = Loader.get_historical_data(
"unknow_file", Source.LocalData, base_date, TimeFrame.Weekly
)