async def getPredictSet(dataset: UploadFile = File(...)):
logging.info("[Received Request]@/predict/dataset")
start = time.time()
try:
res = await dataset.read()
with open(os.path.join(ut.getDatasetsFullPath(dataset.filename)),
"wb") as f:
f.write(res)
r = md.predictFile(ut.getDatasetsFullPath(dataset.filename))
#print(r)
return {
"message": "Success, 成功获取",
'time': time.time() - start,
'filename': dataset.filename,
'result': r
}
except Exception as e:
#print(str(e))
return {
"message": str(e),
'time': time.time() - start,
'filename': dataset.filename
}
def readDataset(dataset):
with open(ut.getDatasetsFullPath(dataset), encoding='utf-8') as d:
return d
pass
import modules.utils as ut
import joblib
import pandas as pd
from sklearn.preprocessing import LabelEncoder
print(ut.getSystemPlatform())
print(ut.getDatasetsFullPath("students.csv"))
print(ut.getModelsFullPath("model.m"))
clf = joblib.load(ut.getModelsFullPath("init_model.m"))
data = pd.read_csv(ut.getDatasetsFullPath("extData.csv"))
# Transform
le = LabelEncoder()
obj_cols = [col for t, col in zip(data.dtypes, data.columns) if t == 'object']
for col in obj_cols:
data[col] = le.fit_transform(data[col])
# Pre-processing
data['All_Sup'] = data['famsup'] & data['schoolsup'] # 1
data['PairEdu'] = data[['Fedu', 'Medu']].mean(axis=1) # 2
data['more_high'] = data['higher'] & (data['schoolsup'] | data['paid']) # 3
data['All_alc'] = data['Walc'] + data['Dalc'] # 4
data['Dalc_per_week'] = data['Dalc'] / data['All_alc'] # 5
data.drop(['Dalc'], axis=1, inplace=True) # 6
data['studytime_ratio'] = data['studytime'] / (
data[['studytime', 'traveltime', 'freetime']].sum(axis=1)) # 7
data.drop(["studytime"], axis=1, inplace=True) # 8
def readDataset(dataset):
return pd.read_csv(ut.getDatasetsFullPath(dataset))