diff --git a/bayes_optim/base.py b/bayes_optim/base.py
index ac56857..dd6ab99 100644
--- a/bayes_optim/base.py
+++ b/bayes_optim/base.py
@@ -39,6 +39,7 @@ def __init__(
         eval_type: str = "list",
         DoE_size: Optional[int] = None,
         warm_data: Tuple = None,
+        initial_guess: np.ndarray = None,
         n_point: int = 1,
         acquisition_fun: str = "EI",
         acquisition_par: dict = None,
@@ -291,7 +292,9 @@ def ask(
         else:  # take the initial DoE
             n_point = self._DoE_size if n_point is None else n_point
             msg = f"asking {n_point} points (using DoE):"
-            X = self.create_DoE(n_point, fixed=fixed)
+            X = self.create_DoE(n_point, fixed=fixed, initial_guess=not self.initial_guess is None)
+            self.logger.info('*** Find me')
+            self.logger.info(X)
 
         if len(X) == 0:
             raise AskEmptyError()
@@ -360,7 +363,7 @@ def tell(
             self.iter_count += 1
             self.hist_f.append(xopt.fitness)
 
-    def create_DoE(self, n_point: int, fixed: Dict = None) -> List:
+    def create_DoE(self, n_point: int, fixed: Dict = None, initial_guess=False) -> List:
         """get the initial sample points using Design of Experiemnt (DoE) methods
 
         Parameters
@@ -377,7 +380,14 @@ def create_DoE(self, n_point: int, fixed: Dict = None) -> List:
         search_space = self.search_space.filter(fixed.keys(), invert=True)
 
         count = 0
-        DoE = []
+        if initial_guess and not self.initial_guess is None:
+            # assert self.initial_guess.shape[0] == self.dim or self.initial_guess.shape[1] == self.dim
+            # assert all([isinstance(_, float) for _ in self.initial_guess[:, self.r_index].ravel()])
+            # assert all([isinstance(_, int) for _ in self.initial_guess[:, self.i_index].ravel()])
+            # assert all([isinstance(_, str) for _ in self.initial_guess[:, self.d_index].ravel()])
+            DoE = [self.initial_guess]
+        else:
+            DoE = []
         while n_point:
             # NOTE: random sampling could generate duplicated points again
             # keep sampling until getting enough points