qis.compute_ewm_cross_xy

qis.compute_ewm_cross_xy(x_data, y_data, span=None, ewm_lambda=0.94, cross_xy_type=CrossXyType.COVAR, mean_adj_type=MeanAdjType.NONE, init_type=InitType.ZERO, var_init_type=InitType.MEAN, nan_backfill=NanBackfill.FFILL)

compute cross ewm for 1-d arrays x and y cross_xy[t] = (1-lambda)*x[t]*y[t] + lambda*cross_xy[t-1]

three supported cases: 1: x and y are pd.DataFrame with same dimensions: z = x*y 2: x is pd.Series, y is pd.DataFrame: ml y=betas*x -> output pandas.dim = pandas.dim y 3: both pd.Series -> output series.dim = series.dim y 4: both np.nd arrays with same dimension

Parameters:

• x_data (DataFrame | Series | ndarray)

• y_data (DataFrame | Series | ndarray)

• span (float | ndarray)

• ewm_lambda (float | ndarray)

• cross_xy_type (CrossXyType)

• mean_adj_type (MeanAdjType)

• init_type (InitType)

• var_init_type (InitType)

• nan_backfill (NanBackfill)

Return type:

DataFrame | Series | ndarray