def main(argv):
conf_file = argv[1]
impstr = "import %s as conf" % conf_file
print impstr
exec impstr
dfAll = None
for each in conf.l_params:
print each
cls = joblib.load(os.path.join(root, 'data', 'models',"model_" + each[0]+ ".pkl"))
sym2ta = model.get_all_from(each[1])
df = model.build_trains(sym2ta, each[3][0], each[3][1])
feat_names = model.get_feat_names(df)
npFeat = df.loc[:,feat_names].values
df["pred"] = cls.predict_proba(npFeat)[:,1]
if dfAll is None:
dfAll = df
else:
dfAll = dfAll.append(df)
dfAll.to_csv(os.path.join(root, 'data', 'crosses', conf_file+ ".csv"))
dacc = accu(dfAll, each[2], 0.0)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"]*1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.5)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"]*1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.6)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"]*1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.7)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"]*1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.8)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"]*1.0 / dacc["pos"]
def main(argv):
conf_file = argv[1]
impstr = "import %s as conf" % conf_file
print impstr
exec impstr
dfAll = None
for each in conf.l_params:
print each
cls = joblib.load(
os.path.join(root, 'data', 'models', "model_" + each[0] + ".pkl"))
sym2ta = model.get_all_from(each[1])
df = model.build_trains(sym2ta, each[3][0], each[3][1])
feat_names = model.get_feat_names(df)
npFeat = df.loc[:, feat_names].values
df["pred"] = cls.predict_proba(npFeat)[:, 1]
if dfAll is None:
dfAll = df
else:
dfAll = dfAll.append(df)
dfAll.to_csv(os.path.join(root, 'data', 'crosses', conf_file + ".csv"))
dacc = accu(dfAll, each[2], 0.0)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"] * 1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.5)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"] * 1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.6)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"] * 1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.7)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"] * 1.0 / dacc["pos"]
dacc = accu(dfAll, each[2], 0.8)
print dacc["trueInPos"], dacc["pos"], dacc["trueInPos"] * 1.0 / dacc["pos"]
def one_work(cls, ta_dir, label, date_range, th):
re = "%s\t%s\t%s\t%s\t%s\t%f\t" % (cls, ta_dir[-4:], label, date_range[0], date_range[1],th)
merged_file = os.path.join(ta_dir, "merged.pkl")
df = filter_(get_df(merged_file))
df = get_range(df, date_range[0], date_range[1])
cls = joblib.load(os.path.join(root, 'data', 'models',"model_" + cls + ".pkl"))
feat_names = model.get_feat_names(df)
npFeat = df.loc[:,feat_names].values
npPred = cls.predict_proba(npFeat)[:,1]
df["pred"] = npPred
dacc = accu(df, label, th)
re += "%d\t%d\t" % (dacc["trueInPos"], dacc["pos"])
if dacc["pos"] > 0:
re += "%f" % (dacc["trueInPos"]*1.0 / dacc["pos"])
else :
re += "0.0"
return re
def one_work(cls, ta_dir, label, date_range, th):
re = "%s\t%s\t%s\t%s\t%s\t%f\t" % (cls, ta_dir[-4:], label, date_range[0],
date_range[1], th)
merged_file = os.path.join(ta_dir, "merged.pkl")
df = filter_(get_df(merged_file))
df = get_range(df, date_range[0], date_range[1])
cls = joblib.load(
os.path.join(root, 'data', 'models', "model_" + cls + ".pkl"))
feat_names = model.get_feat_names(df)
npFeat = df.loc[:, feat_names].values
npPred = cls.predict_proba(npFeat)[:, 1]
df["pred"] = npPred
dacc = accu(df, label, th)
re += "%d\t%d\t" % (dacc["trueInPos"], dacc["pos"])
if dacc["pos"] > 0:
re += "%f" % (dacc["trueInPos"] * 1.0 / dacc["pos"])
else:
re += "0.0"
return re