exo_proc per ts decorator & refactor mech_step

engine/configProcessor.py

@@ -1,5 +1,5 @@
 import pandas as pd
-from functools import partial, reduce
+from functools import reduce
 
 def state_identity(k):
     return lambda step, sL, s, _input: (k, s[k])
@@ -37,9 +37,7 @@ def generate_config(state_dict, mechanisms, exo_proc):
             sdf_values, bdf_values = sdf.values.tolist(), bdf.values.tolist()
         else:
             sdf_values = sdf.values.tolist()
-            # print(sdf_values)
             bdf_values = [ [b_identity] * len(sdf_values) for n in range(mechanisms) ]
-            # print(bdf_values)
         return sdf_values, bdf_values
 
     def only_ep_handler(state_dict):

engine/mechanismExecutor.py

@@ -1,11 +1,14 @@
 from copy import deepcopy
-from fn import op, _
+from fn.op import foldr, call
+from fn import _
+from ui.config import behavior_ops
 
 def getColResults(step, sL, s, funcs):
     return list(map(lambda f: f(step, sL, s), funcs))
 
-def getBehaviorInput(step, sL, s, funcs):
-    return op.foldr(_ + _)(getColResults(step, sL, s, funcs))
+# Data Type reduction
+def getBehaviorInput(step, sL, s, funcs, ops = behavior_ops[::-1]):
+    return foldr(call, getColResults(step, sL, s, funcs))(ops)
 
 def apply_env_proc(env_processes, state_dict, step):
     for state in state_dict.keys():
@@ -21,26 +24,23 @@ def exception_handler(f, m_step, sL, last_mut_obj, _input):
 
 
 def mech_step(m_step, sL, state_funcs, behavior_funcs, env_processes, t_step):
-    in_copy, mutatable_copy, out_copy = deepcopy(sL), deepcopy(sL), deepcopy(sL)
-    last_in_obj, last_mut_obj = in_copy[-1], mutatable_copy[-1]
+    last_in_obj = sL[-1]
 
     _input = exception_handler(getBehaviorInput, m_step, sL, last_in_obj, behavior_funcs)
+    # del last_in_obj
 
-    # print(len(state_funcs))
+    last_in_copy = dict([ exception_handler(f, m_step, sL, last_in_obj, _input) for f in state_funcs ])
+    # print(str(m_step) + ': ' + str(last_in_copy))
 
-    last_mut_obj = dict([
-        exception_handler(f, m_step, sL, last_mut_obj, _input) for f in state_funcs
-    ])
-    # print(str(m_step) + ': ' + str(last_mut_obj))
+    # mutating last_in_copy
+    apply_env_proc(env_processes, last_in_copy, last_in_copy['timestamp'])
 
-    apply_env_proc(env_processes, last_mut_obj, last_mut_obj['timestamp'])
+    last_in_copy["mech_step"], last_in_copy["time_step"] = m_step, t_step
+    sL.append(last_in_copy)
 
-    last_mut_obj["mech_step"], last_mut_obj["time_step"] = m_step, t_step
-    out_copy.append(last_mut_obj)
+    del last_in_copy
 
-    del last_in_obj, last_mut_obj, in_copy, mutatable_copy,
-
-    return out_copy
+    return sL
 
 
 def block_gen(states_list, configs, env_processes, t_step):
@@ -68,12 +68,13 @@ def pipeline(states_list, configs, env_processes, time_seq):
     simulation_list = [states_list]
     for time_step in time_seq:
         pipeline_run = block_gen(simulation_list[-1], configs, env_processes, time_step)
-        head, *pipeline_run = pipeline_run
+        _, *pipeline_run = pipeline_run
         simulation_list.append(pipeline_run)
 
     return simulation_list
 
 
+# Del head
 def simulation(states_list, configs, env_processes, time_seq, runs):
     pipeline_run = []
     for run in range(runs):
@@ -84,7 +85,7 @@ def simulation(states_list, configs, env_processes, time_seq, runs):
             pipeline_run += simulation_list
         else:
             transient_states_list = [pipeline_run[-1][-1]]
-            head, *tail = pipeline(transient_states_list, configs, env_processes, time_seq)
+            _, *tail = pipeline(transient_states_list, configs, env_processes, time_seq)
             pipeline_run += tail
 
     return pipeline_run

engine/run.py

@@ -17,7 +17,7 @@ def main():
     # print(states_list)
     # print(configs)
     # p = pipeline(states_list, configs, env_processes, range(10))
-    T = range(5)
+    T = sim_config['T']
     N = sim_config['N']
 
     # Dimensions: N x r x mechs

engine/utils.py

@@ -1,6 +1,6 @@
 from datetime import datetime, timedelta
 from decimal import Decimal
-from functools import partial
+from fn.func import curried
 
 flatten = lambda l: [item for sublist in l for item in sublist]
 
@@ -35,13 +35,12 @@ def bound_norm_random(rng, low, high):
         res = bound_norm_random(rng, low, high)
     return Decimal(res)
 
-def proc_trigger(trigger_step, update_f):
-    def step_trigger(trigger_step, update_f, step):
-        if step == trigger_step:
-            return update_f
-        else:
-            return lambda x: x
-    return partial(step_trigger, trigger_step, update_f)
+@curried
+def proc_trigger(trigger_step, update_f, step):
+    if step == trigger_step:
+        return update_f
+    else:
+        return lambda x: x
 
 # accept timedelta instead of timedelta params
 def time_step(dt_str, dt_format='%Y-%m-%d %H:%M:%S', days=0, minutes=0, seconds=30):
@@ -56,6 +55,15 @@ def ep_time_step(s, dt_str, fromat_str='%Y-%m-%d %H:%M:%S', days=0, minutes=0, s
     else:
         return dt_str
 
+def exo_update_per_ts(ep):
+    @curried
+    def ep_decorator(f, y, step, sL, s, _input):
+        if s['mech_step'] + 1 == 1:  # inside f body to reduce performance costs
+            return f(step, sL, s, _input)
+        else:
+            return (y, s[y])
+    return {es: ep_decorator(f, es) for es, f in ep.items()}
+
 # def create_tensor_field(mechanisms, env_poc, keys=['behaviors', 'states']):
 #     dfs = [ create_matrix_field(mechanisms, k) for k in keys ]
 #     df = pd.concat(dfs, axis=1)

ui/config.py

@@ -1,4 +1,6 @@
-from engine.utils import bound_norm_random, ep_time_step, proc_trigger
+from engine.utils import bound_norm_random, ep_time_step, proc_trigger, exo_update_per_ts
+from fn.op import foldr
+from fn import _
 
 import numpy as np
 from decimal import Decimal
@@ -58,27 +60,23 @@ def s2m3(step, sL, s, _input):
 # Exogenous States
 proc_one_coef_A = 0.7
 proc_one_coef_B = 1.3
+
 def es3p1(step, sL, s, _input):
     y = 's3'
-    if s['mech_step']+1 == 1:  # inside f body to reduce performance costs
-        x = s['s3'] * bound_norm_random(seed['a'], proc_one_coef_A, proc_one_coef_B)
-        return (y, x)
-    else:
-        return (y, s[y])
+    x = s['s3'] * bound_norm_random(seed['a'], proc_one_coef_A, proc_one_coef_B)
+    return (y, x)
 
 def es4p2(step, sL, s, _input):
     y = 's4'
-    if s['mech_step']+1 == 1: # inside f body to reduce performance costs
-        x = s['s4'] * bound_norm_random(seed['b'], proc_one_coef_A, proc_one_coef_B)
-        return (y, x)
-    else:
-        return (y, s[y])
+    x = s['s4'] * bound_norm_random(seed['b'], proc_one_coef_A, proc_one_coef_B)
+    return (y, x)
 
 def es5p2(step, sL, s, _input): # accept timedelta instead of timedelta params
     y = 'timestamp'
     x = ep_time_step(s, s['timestamp'], seconds=1)
     return (y, x)
 
+
 # Environment States
 def env_a(x):
     return 10
@@ -96,11 +94,13 @@ state_dict = {
     'timestamp': '2018-10-01 15:16:24'
 }
 
-exogenous_states = {
+exogenous_states = exo_update_per_ts(
+    {
     "s3": es3p1,
     "s4": es4p2,
     "timestamp": es5p2
-}
+    }
+)
 
 env_processes = {
     "s3": proc_trigger('2018-10-01 15:16:25', env_a),
@@ -110,6 +110,7 @@ env_processes = {
 # test return vs. non-return functions as lambdas
 # test fully defined functions
 # genesis Sites should always be there
+behavior_ops = [foldr(_ + _), lambda x: x + 0]
 mechanisms = {
     "m1": {
         "behaviors": {
@@ -144,5 +145,6 @@ mechanisms = {
 }
 
 sim_config = {
-    "N": 2
+    "N": 2,
+    "T": range(5)
 }