check_free_disk_space: adapt to statsmodels 0.14.4

Refactor ARIMA calls and correctly specify the minimum version in
setup.py

setup.py

@@ -11,7 +11,7 @@ for f in sorted(glob.glob(os.path.join('slapos', 'README.*.rst'))):
 
 long_description += open("CHANGES.txt").read() + "\n"
 
-prediction_require = ['statsmodels', 'scipy', 'pandas']
+prediction_require = ['statsmodels>=0.14.0', 'scipy', 'pandas']
 test_require = ['mock', 'cryptography', 'websockets; python_version>="3"',] + prediction_require
 
 setup(name=name,

slapos/promise/plugin/check_free_disk_space.py

@@ -22,7 +22,7 @@ from contextlib import closing
 try:
   import pandas as pd
   import numpy as np
-  from statsmodels.tsa.arima_model import ARIMA
+  from statsmodels.tsa.arima.model import ARIMA
 except ImportError:
   pass
 
@@ -127,8 +127,8 @@ class RunPromise(GenericPromise):
       with warnings.catch_warnings():
         warnings.simplefilter("ignore")
         model = ARIMA(history, order=arima_order)
-        model_fit = model.fit(disp=-1)
-        yhat = model_fit.forecast()[0]
+        model_fit = model.fit()
+        yhat = model_fit.get_forecast().predicted_mean[0]
         predictions.append(yhat)
         history.append(test[t])
     # calculate out of sample error
@@ -145,12 +145,12 @@ class RunPromise(GenericPromise):
       for d in d_values:
         for q in q_values:
           order = (p,d,q)
-          try:
-            rmse = self.evaluateArimaModel(dataset, order)
-            if rmse < best_score:
-              best_score, best_cfg = rmse, order
-          except Exception:
-            pass
+          rmse = self.evaluateArimaModel(dataset, order)
+          if rmse < best_score:
+            best_score, best_cfg = rmse, order
+            if rmse == 0.0:
+              self.logger.info("Found perfect model with order %s", order)
+              return best_cfg
     return best_cfg
 
   def diskSpacePrediction(self, disk_partition, database, date, time, day_range):
@@ -191,11 +191,13 @@ class RunPromise(GenericPromise):
           # disabling warnings during the ARIMA calculation
           with warnings.catch_warnings():
             warnings.simplefilter("ignore")
-            model_arima = ARIMA(df, order=best_cfg)
-            # disp < 0 means no output about convergence information
-            model_arima_fit = model_arima.fit(disp=-1)
+            model_arima = ARIMA(df, order=best_cfg, trend="t")
+            model_arima_fit = model_arima.fit()
             # save ARIMA predictions
-            fcast, _, conf = model_arima_fit.forecast(max_date_predicted, alpha=0.05)
+            fcast_result  = model_arima_fit.get_forecast(steps=max_date_predicted)
+            fcast = fcast_result.predicted_mean
+            conf = fcast_result.conf_int(alpha=0.05).to_numpy()
+
           # pass the same index as the others
           fcast = pd.Series(fcast, index=future_index_date)
           if fcast.empty: