Skip to content

Commit

Permalink
Merge pull request #17 from SQRPI/master
Browse files Browse the repository at this point in the history 
Add exclude API(#13)
  • Loading branch information
@SQRPI
SQRPI authored Feb 12, 2020
2 parents 18f00ca + d97b522 commit 71a9cdb
Show file tree
Hide file tree
Showing 2 changed files with 72 additions and 41 deletions.
103 changes: 64 additions & 39 deletions MaterialPlanning.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -7,13 +7,13 @@
headers = {'User-Agent':'ArkPlanner'}

class MaterialPlanning(object):
def __init__(self,

def __init__(self,
filter_freq=200,
filter_stages=[],
url_stats='result/matrix?show_stage_details=true&show_item_details=true',
url_rules='formula',
path_stats='data/matrix.json',
url_rules='formula',
path_stats='data/matrix.json',
path_rules='data/formula.json'):
"""
Object initialization.
Expand All @@ -40,8 +40,8 @@ def __init__(self,
material_probs['matrix'] = filtered_probs

self._set_lp_parameters(*self._pre_processing(material_probs, convertion_rules))


def _pre_processing(self, material_probs, convertion_rules):
"""
Compute costs, convertion rules and items probabilities from requested dictionaries.
Expand All @@ -53,18 +53,18 @@ def _pre_processing(self, material_probs, convertion_rules):
"""
# To count items and stages.
additional_items = {'30135': u'D32钢', '30125': u'双极纳米片', '30115': u'聚合剂'}
exp_unit = 200*30.0/7400
exp_unit = 200*(30.0-0.048*30)/7400
gold_unit = 0.004
exp_worths = {'2001':exp_unit, '2002':exp_unit*2, '2003':exp_unit*5, '2004':exp_unit*10}
gold_worths = {'3003':gold_unit*500}
exp_worths = {'2001':exp_unit, '2002':exp_unit*2, '2003':exp_unit*5, '2004':exp_unit*10, '3003':exp_unit*2}
gold_worths = {}

item_dct = {}
stage_dct = {}
for dct in material_probs['matrix']:
item_dct[dct['item']['itemId']]=dct['item']['name']
stage_dct[dct['stage']['code']]=dct['stage']['code']
item_dct.update(additional_items)

# To construct mapping from id to item names.
item_array = []
item_id_array = []
Expand All @@ -85,7 +85,7 @@ def _pre_processing(self, material_probs, convertion_rules):
stage_array.append(v)
self.stage_array = np.array(stage_array)
self.stage_dct_rv = {v:k for k,v in enumerate(self.stage_array)}

# To format dropping records into sparse probability matrix
probs_matrix = np.zeros([len(stage_array), len(item_array)])
cost_lst = np.zeros(len(stage_array))
Expand All @@ -96,7 +96,8 @@ def _pre_processing(self, material_probs, convertion_rules):
cost_lst[self.stage_dct_rv[dct['stage']['code']]] = dct['stage']['apCost']
float(dct['item']['itemId'])
probs_matrix[self.stage_dct_rv[dct['stage']['code']], self.item_dct_rv[dct['item']['name']]] = dct['quantity']/float(dct['times'])
cost_gold_offset[self.stage_dct_rv[dct['stage']['code']]] = - dct['stage']['apCost']*(12*gold_unit)
if cost_lst[self.stage_dct_rv[dct['stage']['code']]] != 0:
cost_gold_offset[self.stage_dct_rv[dct['stage']['code']]] = - dct['stage']['apCost']*(12*gold_unit)
except:
pass

Expand Down Expand Up @@ -135,40 +136,40 @@ def _pre_processing(self, material_probs, convertion_rules):
for iname in outc_dct:
convertion[self.item_dct_rv[iname]] += outc_dct[iname]*0.175*outc_wgh[iname]/weight_sum
convertion_outc_matrix.append(convertion)

convertion_cost_lst.append(rule['goldCost']*0.004)

convertions_group = (np.array(convertion_matrix), np.array(convertion_outc_matrix), np.array(convertion_cost_lst))
farms_group = (probs_matrix, cost_lst, cost_exp_offset, cost_gold_offset)

return convertions_group, farms_group


def _set_lp_parameters(self, convertions_group, farms_group):
"""
Object initialization.
Args:
convertion_matrix: matrix of shape [n_rules, n_items].
convertion_matrix: matrix of shape [n_rules, n_items].
Each row represent a rule.
convertion_cost_lst: list. Cost in equal value to the currency spent in convertion.
probs_matrix: sparse matrix of shape [n_stages, n_items].
probs_matrix: sparse matrix of shape [n_stages, n_items].
Items per clear (probabilities) at each stage.
cost_lst: list. Costs per clear at each stage.
"""
self.convertion_matrix, self.convertion_outc_matrix, self.convertion_cost_lst = convertions_group
self.probs_matrix, self.cost_lst, self.cost_exp_offset, self.cost_gold_offset = farms_group

assert len(self.probs_matrix)==len(self.cost_lst)
assert len(self.convertion_matrix)==len(self.convertion_cost_lst)
assert self.probs_matrix.shape[1]==self.convertion_matrix.shape[1]
def update(self,


def update(self,
filter_freq=20,
filter_stages=[],
url_stats='result/matrix?show_stage_details=true&show_item_details=true',
url_rules='formula',
path_stats='data/matrix.json',
url_rules='formula',
path_stats='data/matrix.json',
path_rules='data/formula.json'):
"""
To update parameters when probabilities change or new items added.
Expand Down Expand Up @@ -200,16 +201,17 @@ def _get_plan_no_prioties(self, demand_lst, outcome=False, gold_demand=True, exp
Returns:
strategy: list of required clear times for each stage.
fun: estimated total cost.
"""
A_ub = (np.vstack([self.probs_matrix, self.convertion_outc_matrix])
"""
A_ub = (np.vstack([self.probs_matrix, self.convertion_outc_matrix])
if outcome else np.vstack([self.probs_matrix, self.convertion_matrix])).T
farm_cost = (self.cost_lst +
(self.cost_exp_offset if exp_demand else 0) +
farm_cost = (self.cost_lst +
(self.cost_exp_offset if exp_demand else 0) +
(self.cost_gold_offset if gold_demand else 0))
# print(farm_cost)
convertion_cost_lst = self.convertion_cost_lst if gold_demand else np.zeros(self.convertion_cost_lst.shape)
cost = (np.hstack([farm_cost, convertion_cost_lst]))
assert np.any(farm_cost>=0)

excp_factor = 1.0
dual_factor = 1.0

Expand All @@ -220,7 +222,7 @@ def _get_plan_no_prioties(self, demand_lst, outcome=False, gold_demand=True, exp
method='interior-point')
if solution.status != 4:
break

excp_factor /= 10.0

while dual_factor>1e-5:
Expand All @@ -232,13 +234,13 @@ def _get_plan_no_prioties(self, demand_lst, outcome=False, gold_demand=True, exp
break

dual_factor /= 10.0


return solution, dual_solution, excp_factor


def get_plan(self, requirement_dct, deposited_dct={},
print_output=True, outcome=False, gold_demand=True, exp_demand=True):
def get_plan(self, requirement_dct, deposited_dct={},
print_output=True, outcome=False, gold_demand=True, exp_demand=True, exclude=[]):
"""
User API. Computing the material plan given requirements and owned items.
Args:
Expand All @@ -256,7 +258,23 @@ def get_plan(self, requirement_dct, deposited_dct={},
demand_lst[self.item_dct_rv[k]] = v
for k, v in deposited_dct.items():
demand_lst[self.item_dct_rv[k]] -= v


is_stage_alive = [False if stage in exclude else True for stage in self.stage_array]

if exclude:
BackTrace = [
copy.copy(self.stage_array),
copy.copy(self.cost_lst),
copy.copy(self.probs_matrix),
copy.copy(self.cost_exp_offset),
copy.copy(self.cost_gold_offset)
]
self.stage_array = self.stage_array[is_stage_alive]
self.cost_lst = self.cost_lst[is_stage_alive]
self.probs_matrix = self.probs_matrix[is_stage_alive]
self.cost_exp_offset = self.cost_exp_offset[is_stage_alive]
self.cost_gold_offset = self.cost_gold_offset[is_stage_alive]

stt = time.time()
solution, dual_solution, excp_factor = self._get_plan_no_prioties(demand_lst, outcome, gold_demand, exp_demand)
x, status = solution.x/excp_factor, solution.status
Expand Down Expand Up @@ -321,7 +339,7 @@ def get_plan(self, requirement_dct, deposited_dct={},
}
values[int(self.item_id_array[i][-1])-1]['items'].append(item_value)
for group in values:
group["items"] = sorted(group["items"], key=lambda k: float(k['value']), reverse=True)
group["items"] = sorted(group["items"], key=lambda k: float(k['value']), reverse=True)

res = {
"cost": int(cost),
Expand Down Expand Up @@ -353,6 +371,13 @@ def get_plan(self, requirement_dct, deposited_dct={},
print('Level %d items: '%(i+1))
print(', '.join(display_lst))

if exclude:
self.stage_array = BackTrace[0]
self.cost_lst = BackTrace[1]
self.probs_matrix = BackTrace[2]
self.cost_exp_offset = BackTrace[3]
self.cost_gold_offset = BackTrace[4]

return res


Expand Down Expand Up @@ -391,7 +416,7 @@ def request_data(url_stats, url_rules, save_path_stats, save_path_rules):
os.mkdir(os.path.dirname(save_path_rules))
except:
pass

req = urllib.request.Request(url_stats, None, headers)
with urllib.request.urlopen(req) as response:
material_probs = json.loads(response.read().decode())
Expand All @@ -417,9 +442,9 @@ def load_data(path_stats, path_rules):
material_probs: dictionary. Content of the stats json file.
convertion_rules: dictionary. Content of the rules json file.
"""
with open(path_stats) as json_file:
with open(path_stats) as json_file:
material_probs = json.load(json_file)
with open(path_rules) as json_file:
with open(path_rules) as json_file:
convertion_rules = json.load(json_file)

return material_probs, convertion_rules
10 changes: 8 additions & 2 deletions server.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -40,12 +40,18 @@ async def plan(request):
except:
gold_demand = True

try:
exclude = request.json['exclude']
except:
exclude = []

try:
if time.time() - last_updated > 60 * 30:
mp.update()
last_updated = time.time()
dct = mp.get_plan(required_dct, owned_dct, False,
outcome=extra_outc, exp_demand=exp_demand, gold_demand=gold_demand)
dct = mp.get_plan(required_dct, owned_dct, False,
outcome=extra_outc, exp_demand=exp_demand, gold_demand=gold_demand,
exclude=exclude)
except ValueError as e:
return response.json({"error": True, "reason": str(e)})

Expand Down

0 comments on commit 71a9cdb

Please sign in to comment.