def merge_pyalgo_accdb(pyalgo, accdb):
pyalgo = pyalgo.copy()
pyalgo.columns = wio.flatten_columns(pyalgo.columns)
merged = pd.concat((pyalgo, accdb), axis=1, keys=("pyalgo", "accdb"))
merged.columns = wio.flatten_columns(merged.columns)
return merged
def test_flatten_columns():
cols = pd.MultiIndex.from_tuples([("a", "aa"), ("b", "")],
names=("gear", "item"))
fcols = wio.flatten_columns(cols)
infcols = wio.inflate_columns(fcols)
assert cols.equals(infcols)
assert cols.names == infcols.names
with pytest.raises(AssertionError, match="MultiIndex?"):
wio.inflate_columns(cols)
def is_more_low_powered_gears(cyc):
cyc2 = cyc.copy()
cyc2.columns = wio.inflate_columns(cyc2.columns)
c22 = cyc2.iloc[1571:1579][["ok_p", "ok_max_n"]].dropna(axis=1, how="all")
try:
## is there any row with all-low-p AND 2-or-more n-max-ok?
bad_rows = (~c22["ok_p"].replace(-1, 0).astype("bool")).all(axis=1) & (
c22["ok_max_n"].replace(-1, 0).astype("bool").sum(axis=1) > 1)
if bad_rows.any():
c22.columns = wio.flatten_columns(c22.columns)
return (
list(bad_rows[bad_rows].index),
pd.concat(
(c22, cyc.iloc[1571:1579].loc[:, ["g_min", "g_max"]]),
axis=1),
)
except Exception as ex:
print(ex)
def display_pyalgo_flags(pyalgo, accdb, props):
flags = pyalgo.select_dtypes("int8").copy()
flags.columns = wio.flatten_columns(flags.columns)
pyalgo = pyalgo[["p_req", "p_avail"]].copy()
pyalgo.columns = wio.flatten_columns(pyalgo.columns)
grid(pd.concat((pyalgo, flags), axis=1))
def display_pyalgo(pyalgo, accdb, props):
pyalgo.columns = wio.flatten_columns(pyalgo.columns)
grid(pyalgo.drop("t", axis=1))
# veh_nums += [82, 88, 91, 99, 100, 101, 112, 113, 114]
# # UNCOMMENT next line to FETCH new vehicles.
p1, c1, p2, c2 = load_accdb_and_python_datasets()
equivalent_series = [
("v_orig", "V_cycle"),
("v_downscale", "v_target"),
# ("a", "a_target"),
("P_tot_set", "p_req"),
# ("P_max", "p_available"),
("g_max", "g_max0"),
("g_min", "g_min"),
# ("gear", "gears"),
# ("nc", "n"),
]
c2.columns = wio.flatten_columns(c2.columns)
cols1, cols2 = zip(*equivalent_series)
cols1, cols2 = list(cols1), list(cols2)
## Concat props to col-aggregates & convert prop-cols to numerics
cc1 = pd.concat((p1.infer_objects(), c1[cols1].abs().mean(level=0)), axis=1)
cc2 = pd.concat((p2.infer_objects(), c2[cols2].abs().mean(level=0)), axis=1)
equivalent_props = [
("Description", None),
("vehicle_class", "wltc_class"),
("pmr_km", None),
("no_of_gears", None),
("f_dsc_req", "f_dsc"),
("v_max", "v_max"),
("n_vmax", "n_vmax"),
