def gen_user_get_shop_coupon(offline_source, online_source, X):
offline = Matrix(
np.genfromtxt(paths.ccf_path + offline_source,
delimiter=',',
dtype=str),
["uid", "mid", "cid", "dis_rate", "dist", "date_rec", "date"],
["%s" for i in xrange(7)])
online = Matrix(
np.genfromtxt(paths.ccf_path + online_source, delimiter=',',
dtype=str),
["uid", "mid", "act", "cid", "dis_rate", "date_rec", "date"],
["%s" for i in xrange(7)])
user_get_shop_coupon = HashSet()
for i in xrange(offline.ndata):
uid_str = offline.get_cell(i, "uid")
cid_str = offline.get_cell(i, "cid")
mid_str = offline.get_cell(i, "mid")
if cid_str != 'null':
if not user_get_shop_coupon.has(uid_str):
user_get_shop_coupon.set(uid_str, HashSet())
user_get_shop_coupon.get(uid_str).add_one(mid_str)
for i in xrange(online.ndata):
uid_str = online.get_cell(i, "uid")
cid_str = online.get_cell(i, "cid")
mid_str = online.get_cell(i, "mid")
act_str = online.get_cell(i, "act")
if cid_str != 'null' and act_str != '0':
if not user_get_shop_coupon.has(uid_str):
user_get_shop_coupon.set(uid_str, HashSet())
user_get_shop_coupon.get(uid_str).add_one(mid_str)
X.join_by_double_key("uid",
"mid",
"user_get_shop_coupon",
user_get_shop_coupon,
"%s",
dft=0.0)
def divide(x, y):
assert float(x) <= float(y)
if float(y) == 0:
return 0.0
else:
return float(x) * 1.0 / float(y)
X.gen_arith_feature("user_buy_with_coupon_in_shop",
"user_get_shop_coupon",
"user_use_shop_coupon_freq",
divide,
"%s",
dft=0.0)
def gen_user_buy_in_shop(matrix_offline, matrix_online, X):
user_buy_in_shop = HashSet(default=matrix_offline.default)
for i in xrange(matrix_offline.ndata):
date_str = matrix_offline.get_cell(i, 'date')
uid_str = matrix_offline.get_cell(i, 'uid')
mid_str = matrix_offline.get_cell(i, "mid")
if date_str != 'null':
if not user_buy_in_shop.has(uid_str):
user_buy_in_shop.set(uid_str, HashSet())
user_buy_in_shop.get(uid_str).add_one(mid_str)
for i in xrange(matrix_online.ndata):
act_str = matrix_online.get_cell(i, 'act')
uid_str = matrix_online.get_cell(i, 'uid')
mid_str = matrix_online.get_cell(i, "mid")
if act_str == '1':
if not user_buy_in_shop.has(uid_str):
user_buy_in_shop.set(uid_str, HashSet())
user_buy_in_shop.get(uid_str).add_one(mid_str)
X.join_by_double_key("uid",
"mid",
"user_buy_in_shop",
user_buy_in_shop,
"%s",
dft=0.0)
def divide(x, y):
assert float(x) <= float(y)
if float(y) == 0:
return 0
else:
return float(x) * 1.0 / float(y)
X.gen_arith_feature("user_buy_in_shop",
"user_buy",
"user_buy_in_shop_ratio",
divide,
fmt="%s",
dft=0.0)
def gen_merchant_share(X, month):
offline = Matrix(
np.genfromtxt(paths.ccf_path + 'offline_train_{0}.csv'.format(month),
delimiter=',',
dtype=str),
["uid", "mid", "cid", "dis_rate", "dist", "date_rec", "date"],
["%s" for i in xrange(7)])
online = Matrix(
np.genfromtxt(paths.ccf_path + 'online_train_{0}.csv'.format(month),
delimiter=',',
dtype=str),
["uid", "mid", "act", "cid", "dis_rate", "date_rec", "date"],
["%s" for i in xrange(7)])
merchant_user_buy = HashSet()
merchant_user_use_coupon = HashSet()
merchant_user_buy_counter = HashSet()
merchant_user_use_coupon_counter = HashSet()
for i in xrange(offline.ndata):
if i % 100000 == 0:
print i
mid_str = offline.get_cell(i, "mid")
uid_str = offline.get_cell(i, "uid")
date_str = offline.get_cell(i, "date")
cid_str = offline.get_cell(i, "cid")
if date_str != 'null':
if not merchant_user_buy.has(mid_str):
merchant_user_buy.set(mid_str, HashSet())
if not merchant_user_buy.get(mid_str).has(uid_str):
merchant_user_buy.get(mid_str).add_one(uid_str)
merchant_user_buy_counter.add_one(mid_str)
if cid_str != 'null':
if not merchant_user_use_coupon.has(mid_str):
merchant_user_use_coupon.set(mid_str, HashSet())
if not merchant_user_use_coupon.get(mid_str).has(uid_str):
merchant_user_use_coupon.get(mid_str).add_one(uid_str)
merchant_user_use_coupon_counter.add_one(mid_str)
for i in xrange(online.ndata):
if i % 100000 == 0:
print i
mid_str = online.get_cell(i, "mid")
uid_str = online.get_cell(i, "uid")
act_str = online.get_cell(i, "act")
cid_str = online.get_cell(i, "cid")
if act_str == '1':
if not merchant_user_buy.has(mid_str):
merchant_user_buy.set(mid_str, HashSet())
if not merchant_user_buy.get(mid_str).has(uid_str):
merchant_user_buy.get(mid_str).add_one(uid_str)
merchant_user_buy_counter.add_one(mid_str)
if cid_str != 'null':
if not merchant_user_use_coupon.has(mid_str):
merchant_user_use_coupon.set(mid_str, HashSet())
if not merchant_user_use_coupon.get(mid_str).has(uid_str):
merchant_user_use_coupon.get(mid_str).add_one(uid_str)
merchant_user_use_coupon_counter.add_one(mid_str)
X.join("mid", ["merchant_user_buy"],
merchant_user_buy_counter, ["%s"],
dft=0.0)
X.join("mid", ["merchant_user_use_coupon"],
merchant_user_use_coupon_counter, ["%s"],
dft=0.0)
X.check_point(month)
def gen_user_buy_coupon_in_shop(matrix_offline, matrix_online, X):
user_buy_with_coupon_in_shop = HashSet()
user_buy_without_coupon_in_shop = HashSet()
for i in xrange(matrix_offline.ndata):
uid_str = matrix_offline.get_cell(i, "uid")
cid_str = matrix_offline.get_cell(i, "cid")
date_str = matrix_offline.get_cell(i, "date")
mid_str = matrix_offline.get_cell(i, "mid")
if date_str != 'null':
if not user_buy_with_coupon_in_shop.has(uid_str):
user_buy_with_coupon_in_shop.set(uid_str, HashSet())
user_buy_without_coupon_in_shop.set(uid_str, HashSet())
if cid_str != 'null':
user_buy_with_coupon_in_shop.get(uid_str).add_one(mid_str)
else:
user_buy_without_coupon_in_shop.get(uid_str).add_one(mid_str)
for i in xrange(matrix_online.ndata):
uid_str = matrix_online.get_cell(i, "uid")
cid_str = matrix_online.get_cell(i, "cid")
mid_str = matrix_online.get_cell(i, "mid")
act_str = matrix_online.get_cell(i, "act")
if act_str == '1':
if not user_buy_with_coupon_in_shop.has(uid_str):
user_buy_with_coupon_in_shop.set(uid_str, HashSet())
user_buy_without_coupon_in_shop.set(uid_str, HashSet())
if cid_str != 'null':
user_buy_with_coupon_in_shop.get(uid_str).add_one(mid_str)
else:
user_buy_without_coupon_in_shop.get(uid_str).add_one(mid_str)
X.join_by_double_key("uid",
"mid",
"user_buy_with_coupon_in_shop",
user_buy_with_coupon_in_shop,
"%s",
dft=0.0)
X.join_by_double_key("uid",
"mid",
"user_buy_without_coupon_in_shop",
user_buy_without_coupon_in_shop,
"%s",
dft=0.0)
def divide(x, y):
assert float(x) <= float(y)
if float(y) == 0:
return 0
else:
return float(x) * 1.0 / float(y)
X.gen_arith_feature("user_buy_with_coupon_in_shop",
"user_buy_in_shop",
"user_buy_with_coupon_in_shop_ratio",
divide,
"%s",
dft=0.0)
X.gen_arith_feature("user_buy_without_coupon_in_shop",
"user_buy_in_shop",
"user_buy_without_coupon_in_shop_ratio",
divide,
"%s",
dft=0.0)
def gen_coupon_gap(X, month):
get_coupon_history = HashSet()
get_coupon_history_user = HashSet()
for i in xrange(X.ndata):
if i % 100000 == 0:
print i
uid_str = X.get_cell(i, "uid")
mid_str = X.get_cell(i, "mid")
date_rec_str = X.get_cell(i, "date_rec")
if date_rec_str != 'null':
history_list = get_coupon_history.get(uid_str,
HashSet()).get(mid_str, [])
#if date_rec_str not in history_list:
history_list.append(date_rec_str)
user_history_list = get_coupon_history_user.get(uid_str, [])
#if date_rec_str not in user_history_list:
user_history_list.append(date_rec_str)
if month > 1:
last_X = Matrix(
np.genfromtxt(paths.ccf_path +
'offline_1_month{0}.csv'.format(month - 1),
delimiter=',',
dtype=str),
["uid", "mid", "cid", "dis_rate", "dist", "date_rec", "date"],
["%s" for i in xrange(7)])
for i in xrange(last_X.ndata):
if i % 100000 == 0:
print i
uid_str = last_X.get_cell(i, "uid")
mid_str = last_X.get_cell(i, "mid")
date_rec_str = last_X.get_cell(i, "date_rec")
if date_rec_str != 'null':
history_list = get_coupon_history.get(uid_str, HashSet()).get(
mid_str, [])
#if date_rec_str not in history_list:
history_list.append(date_rec_str)
user_history_list = get_coupon_history_user.get(uid_str, [])
#if date_rec_str not in user_history_list:
user_history_list.append(date_rec_str)
if month < 7:
next_X = Matrix(
np.genfromtxt(paths.ccf_path +
'offline_1_month{0}.csv'.format(month + 1),
delimiter=',',
dtype=str),
["uid", "mid", "cid", "dis_rate", "dist", "date_rec", "date"],
["%s" for i in xrange(7)])
for i in xrange(next_X.ndata):
if i % 100000 == 0:
print i
uid_str = next_X.get_cell(i, "uid")
mid_str = next_X.get_cell(i, "mid")
date_rec_str = next_X.get_cell(i, "date_rec")
if date_rec_str != 'null':
history_list = get_coupon_history.get(uid_str, HashSet()).get(
mid_str, [])
#if date_rec_str not in history_list:
history_list.append(date_rec_str)
user_history_list = get_coupon_history_user.get(uid_str, [])
#if date_rec_str not in user_history_list:
user_history_list.append(date_rec_str)
for i in xrange(X.ndata):
if i % 100000 == 0:
print i
uid_str = X.get_cell(i, "uid")
mid_str = X.get_cell(i, "mid")
history_list = get_coupon_history.get(uid_str,
HashSet()).get(mid_str, [])
get_coupon_history.get(uid_str).set(mid_str, np.sort(history_list))
user_history_list = get_coupon_history_user.get(uid_str, [])
get_coupon_history_user.set(uid_str, np.sort(user_history_list))
col_names = [
"prev_gap", "next_gap", "user_prev_gap", "user_next_gap", "in_between",
"prev_gap_prev", "next_gap_prev"
]
gaps = np.zeros((X.ndata, len(col_names)))
for i in xrange(X.ndata):
if i % 100000 == 0:
print i
uid_str = X.get_cell(i, "uid")
mid_str = X.get_cell(i, "mid")
date_rec_str = X.get_cell(i, "date_rec")
history_list = get_coupon_history.get(uid_str).get(mid_str)
user_history_list = get_coupon_history_user.get(uid_str)
k = -1
for j in xrange(len(history_list)):
##if history_list[j] == date_rec_str:
if j == len(history_list) - 1 or history_list[j +
1] > date_rec_str:
break
else:
k += 1
in_between = 1
if k > 0:
prev_gap = days_dis(history_list[k], date_rec_str)
else:
prev_gap = 0
in_between = 0
if k + 2 < len(history_list):
next_gap = days_dis(date_rec_str, history_list[k + 2])
else:
next_gap = 0
in_between = 0
k = -1
for j in xrange(len(user_history_list)):
##if user_history_list[j] == date_rec_str:
if j == len(history_list) - 1 or history_list[j +
1] > date_rec_str:
break
else:
k += 1
if k > 0:
user_prev_gap = days_dis(user_history_list[k], date_rec_str)
else:
user_prev_gap = 0
if k + 2 < len(user_history_list):
user_next_gap = days_dis(date_rec_str, user_history_list[k + 2])
else:
user_next_gap = 0
k = -1
for j in xrange(len(history_list)):
if history_list[j] == date_rec_str:
##if j == len(history_list) - 1 or history_list[j + 1] > date_rec_str:
break
else:
k += 1
if k > 0:
prev_gap_prev = days_dis(history_list[k], date_rec_str)
else:
prev_gap_prev = 0
if k + 2 < len(history_list):
next_gap_prev = days_dis(date_rec_str, history_list[k + 2])
else:
next_gap_prev = 0
in_between = 0
gaps[i, :] = np.array([
prev_gap, next_gap, user_prev_gap, user_next_gap, in_between,
prev_gap_prev, next_gap_prev
])
X.cat_col(gaps, col_names, ["%s" for i in xrange(len(col_names))])
X.check_point("X_{0}_{1}{2}_gap".format(month, 1, month - 1))
def gen_user_shop_month(X, month):
user_shop_month = HashSet()
shop_month_coupon = {}
shop_month_coupon_used = {}
shop_coupon = {}
shop_coupon_used = {}
for i in xrange(X.ndata):
if i % 100000 == 0:
print i
uid_str = X.get_cell(i, "uid")
mid_str = X.get_cell(i, "mid")
date_rec_str = X.get_cell(i, "date_rec")
if date_rec_str == 'null':
continue
month = get_month(date_rec_str)
user_shop_month_set = user_shop_month.get(uid_str,
HashSet()).get(mid_str, {})
if month not in user_shop_month_set:
user_shop_month_set[month] = 1
else:
user_shop_month_set[month] += 1
"""
for i in xrange(X.ndata):
if i % 100000 == 0:
print i
uid_str = X.get_cell(i, "uid")
mid_str = X.get_cell(i, "mid")
date_rec_str = X.get_cell(i, "date_rec")
date_str = X.get_cell(i, "date")
cid_str = X.get_cell(i, "cid")
if date_rec_str == 'null':
continue
assert cid_str != 'null'
month = get_month(date_rec_str)
user_shop_month_set = user_shop_month.get(uid_str, HashSet()).get(mid_str, {})
if mid_str not in shop_coupon:
shop_coupon[mid_str] = 1
else:
shop_coupon[mid_str] += 1
if date_str != 'null':
if mid_str not in shop_coupon_used:
shop_coupon_used[mid_str] = 1
else:
shop_coupon_used[mid_str] += 1
if month in user_shop_month_set and user_shop_month_set[month] > 1:
if mid_str not in shop_month_coupon:
shop_month_coupon[mid_str] = 1
else:
shop_month_coupon[mid_str] += 1
if date_str != 'null':
if mid_str not in shop_month_coupon_used:
shop_month_coupon_used[mid_str] = 1
else:
shop_month_coupon_used[mid_str] += 1
"""
#col_names = ["user_shop_month_11_2", "shop_coupon_11_2", "shop_coupon_used_11_2", "shop_month_coupon_11_2", "shop_month_coupon_used_11_2", "shop_coupon_used_ratio_11_2", "is_shop_coupon_11_2"]
col_names = ["user_shop_month_11_2", "is_shop_coupon_11_2"]
statistics = np.zeros((X.ndata, len(col_names)), dtype=float)
for i in xrange(X.ndata):
if i % 100000 == 0:
print i
uid_str = X.get_cell(i, "uid")
mid_str = X.get_cell(i, "mid")
user_shop_month_set = user_shop_month.get(uid_str,
HashSet()).get(mid_str, {})
#if shop_month_coupon.get(mid_str, 0) == 0:
# ratio = 0
#else:
# ratio = shop_month_coupon_used.get(mid_str, 0) * 1.0 / shop_month_coupon.get(mid_str)
if user_shop_month_set.get(month, 0) > 1:
is_shop_coupon = 1
else:
is_shop_coupon = 0
statistics[i, :] = np.array(
[user_shop_month_set.get(month, 0), is_shop_coupon])
X.cat_col(statistics, col_names, ["%s" for i in xrange(len(col_names))])
X.check_point("X_{0}_{1}{2}_shop_month".format(month, 1, month - 1))