Add ARIMA Normalization Functionality

[ml874]

ARIMA here is used without the moving averages component to normalize and forecast time series data.

An ARIMA model is selected from 9 possible combinations: (0,0,0), (1,0,0), (2,0,0), (0,1,0), (1,1,0), (2,1,0), (0,2,0), (1,2,0), (2,2,0). The time series is split into train and test sets and an ARIMA model is fit for every combination on the training set. The model with the lowest mean-squared error (MSE) on the test set is selected as the best model. The original times series can then be transformed by the best model.

import pandas as pd
import numpy as np
import gs_quant.timeseries as ts

series = pd.Series(np.random.randint(0,100,size=(1, 100))[0])

arima = ts.arima()
arima.fit(series, train_size=0.8)

arima.transform(series)

[AnastasiyaB]

the file should live in one of the packages - either statistics or econometrics

[AnastasiyaB]

train size should be an int

[ml874]

I changed it so it could take in a float, int or None (similar to what scikit-learn does).

• If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the train split. If int, represents the absolute number of train samples. If None, the value is automatically set 0.75 (https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html).

Is that too complicated or should we just use int to simplify things

[AnastasiyaB]

pls raise exception, remove print

[ml874]

Certain combinations of (p, q, d) will raise the following exception: Estimation requires the inclusion of least one AR term, MA term, a constant or an exogenous variable.

Raise exception will then break the training loop. Maybe it's a better idea to just print the error and move on to the next combination of (p, q, d)?