Observation get forecast from dict instread of action to fix issue #131

Use private attribute _forecasted_inj to get values instead of using the action built from _forecasted_inj in simulate; which removes the need to call simulate

CHANGELOG.rst

@@ -22,6 +22,7 @@ Change Log
 - [FIXED] `Issue #129<https://github.com/rte-france/Grid2Op/issues/129>`_: game over count for env_actions
 - [FIXED] `Issue #127 <https://github.com/rte-france/Grid2Op/issues/127>`_: Removed no longer existing attribute docstring `indisponibility`
 - [FIXED] `Issue #133 <https://github.com/rte-france/Grid2Op/issues/133>`_: Missing positional argument `space_prng` in `Action.SerializableActionSpace`
+- [FIXED] `Issue #131 <https://github.com/rte-france/Grid2Op/issues/131>`_: Forecast values are accessible without needing to call `obs.simulate` beforehand.
 
 [1.1.1] - 2020-07-07
 ---------------------

grid2op/Observation/BaseObservation.py

@@ -661,20 +661,20 @@ def get_forecasted_inj(self, time_step=1):
         """
         if time_step >= len(self._forecasted_inj):
             raise NoForecastAvailable("Forecast for {} timestep ahead is not possible with your chronics.".format(time_step))
-        a = self._forecasted_grid_act[time_step]["inj_action"]
+        t, a = self._forecasted_inj[time_step]
         prod_p_f = np.full(self.n_gen, fill_value=np.NaN, dtype=dt_float)
         prod_v_f = np.full(self.n_gen, fill_value=np.NaN, dtype=dt_float)
         load_p_f = np.full(self.n_load, fill_value=np.NaN, dtype=dt_float)
         load_q_f = np.full(self.n_load, fill_value=np.NaN, dtype=dt_float)
 
-        if "prod_p" in a._dict_inj:
-            prod_p_f = a._dict_inj["prod_p"]
-        if "prod_v" in a._dict_inj:
-            prod_v_f = a._dict_inj["prod_v"]
-        if "load_p" in a._dict_inj:
-            load_p_f = a._dict_inj["load_p"]
-        if "load_q" in a._dict_inj:
-            load_q_f = a._dict_inj["load_q"]
+        if "prod_p" in a["injection"]:
+            prod_p_f = a["injection"]["prod_p"]
+        if "prod_v" in a["injection"]:
+            prod_v_f = a["injection"]["prod_v"]
+        if "load_p" in a["injection"]:
+            load_p_f = a["injection"]["load_p"]
+        if "load_q" in a["injection"]:
+            load_q_f = a["injection"]["load_q"]
         tmp_arg = ~np.isfinite(prod_p_f)
         prod_p_f[tmp_arg] = self.prod_p[tmp_arg]
         tmp_arg = ~np.isfinite(prod_v_f)