adding necessary functions to time_iteration docstring

doc/source/algos.rst

@@ -1,6 +1,24 @@
 Solution algorithms
 *******************
 
+Solution algorithms operate on model objects to determine
+the optimal decision rules for the model agents.
+
+Each algorithm requires its target model to have functions
+defined in its specification.
+
+This is a mapping between solution algorithm and necessary functions.
+Description of these functions, including their type signature,
+can be found in the Model Specification.
+
+- Improved time iteration: `transition` and `arbitrage`
+- Parameterized expectations : `transition`, `expectation`, and `direct response`
+- Perfect foresight: `transition` and `arbitrage`
+- Perturbation : `transition` and `arbitrage`
+- Steady-state : `transition` and `arbitrage`
+- Time iteration: `transition` and `arbitrage`
+- Value function iteration : `transition`, and either `felicity` or `value`
+
 .. toctree::
     :maxdepth: 2
 

dolo/algos/time_iteration.py

@@ -44,7 +44,8 @@ def time_iteration(model, dr0=None, with_complementarities=True,
     Parameters
     ----------
     model : Model
-        model to be solved
+        model to be solved.
+        Must have ``transition`` and ``arbitrage`` functions.
     verbose : boolean
         if True, display iterations
     dr0 : decision rule