Merge pull request #1 from EC-AI/main

Option of using (neqs,) and other changes and cleaning

statsmodels/tsa/vector_ar/util.py

@@ -4,6 +4,7 @@
 """
 from statsmodels.compat.pandas import frequencies
 from statsmodels.compat.python import asbytes
+from statsmodels.tools.validation import array_like
 
 import numpy as np
 import pandas as pd
@@ -190,7 +191,7 @@ def acf_to_acorr(acf):
     return acf / np.sqrt(np.outer(diag, diag))
 
 
-def varsim(coefs, intercept, sig_u, steps=100, initvalues=None, seed=None):
+def varsim(coefs, intercept, sig_u, steps=100, initial_values=None, seed=None):
     """
     Simulate VAR(p) process, given coefficients and assuming Gaussian noise
 
@@ -215,11 +216,16 @@ def varsim(coefs, intercept, sig_u, steps=100, initvalues=None, seed=None):
         observations to start the autoregressive process.
         If offset is not None, then exog of the model are used if they were
         provided in the model
-    initvalues : None or ndarray (nlags, neqs)
-        Initial values for use in the simulation.
+    initial_values : array_like, optional
+        Initial values for use in the simulation. Shape should be 
+        (nlags, neqs) or (neqs,). Values should be ordered from less to
+        most recent. Note that this values will be returned by the 
+        simulation as the first values of `endog_simulated` and they 
+        will count for the total number of steps.
     seed : {None, int}
         If seed is not None, then it will be used with for the random
         variables generated by numpy.random.
+
     Returns
     -------
     endog_simulated : nd_array
@@ -242,11 +248,12 @@ def varsim(coefs, intercept, sig_u, steps=100, initvalues=None, seed=None):
         result[p:] += ugen[p:]
     else:
         result[p:] = ugen[p:]
-
-    if initvalues is not None:
-        if not initvalues.shape == (p, k):
-            raise ValueError("Initial values should have shape (p, k) where p is the number of lags and k is the number of equations.")
-        result[:p] = initvalues
+
+    initial_values = array_like(initial_values, "initial_values", optional=True, maxdim=2)
+    if initial_values is not None:
+        if not (initial_values.shape == (p, k) or initial_values.shape == (k,)):
+            raise ValueError("initial_values should have shape (p, k) or (k,) where p is the number of lags and k is the number of equations.")
+        result[:p] = initial_values
 
     # add in AR terms
     for t in range(p, steps):

statsmodels/tsa/vector_ar/var_model.py

@@ -923,7 +923,7 @@ def is_stable(self, verbose=False):
         """
         return is_stable(self.coefs, verbose=verbose)
 
-    def simulate_var(self, steps=None, offset=None, initial_state=None, seed=None):
+    def simulate_var(self, steps=None, offset=None, seed=None, initial_values=None):
         """
         simulate the VAR(p) process for the desired number of steps
 
@@ -941,11 +941,15 @@ def simulate_var(self, steps=None, offset=None, initial_state=None, seed=None):
             the linear predictor of those components will be used as offset.
             This should have the same number of rows as steps, and the same
             number of columns as endogenous variables (neqs).
-        initial_state : None or ndarray (nlags, neqs)
-            Initial values for use in the simulation.
         seed : {None, int}
             If seed is not None, then it will be used with for the random
             variables generated by numpy.random.
+        initial_values : array_like, optional
+            Initial values for use in the simulation. Shape should be 
+            (nlags, neqs) or (neqs,). Values should be ordered from less to
+            most recent. Note that this values will be returned by the 
+            simulation as the first values of `endog_simulated` and they 
+            will count for the total number of steps.
 
         Returns
         -------
@@ -981,7 +985,7 @@ def simulate_var(self, steps=None, offset=None, initial_state=None, seed=None):
                 )
 
         y = util.varsim(
-            self.coefs, offset, self.sigma_u, steps=steps, initvalues=initial_state, seed=seed
+            self.coefs, offset, self.sigma_u, steps=steps, seed=seed, initial_values=initial_values, 
         )
         return y