美购精排模型

mlp/train.py

@@ -39,18 +39,18 @@ def getDataVocabFromRedis(version):
     return dataVocab
 
 # 数据类型转换
-def csvTypeConvert(df,data_vocab):
-    # 离散na值填充
-    for k, v in data_vocab.items():
-        df[k] = df[k].fillna("-1")
-        df[k] = df[k].astype("string")
-
-    for k in ITEM_NUMBER_COLUMNS:
-        df[k] = df[k].fillna(0.0)
-        df[k] = df[k].astype("float")
+def csvTypeConvert(columns,df,data_vocab):
+    for k in columns:
+        # 离散na值填充
+        if k in data_vocab.items():
+            df[k] = df[k].astype("string")
+            df[k] = df[k].fillna("-1")
+        else:
+            df[k] = df[k].astype("float")
+            df[k] = df[k].fillna(0.0)
 
     df["label"] = df["label"].astype("int")
-    print(df.dtypes)
+    # print(df.dtypes)
     return df
 
 def loadData(data_path):
@@ -166,8 +166,8 @@ if __name__ == '__main__':
     print("trainSize:{},testSize{}".format(trainSize,testSize))
 
     # 数据类型转换
-    df_train = csvTypeConvert(df_train,data_vocab)
-    df_test = csvTypeConvert(df_test,data_vocab)
+    df_train = csvTypeConvert(datasColumns,df_train,data_vocab)
+    df_test = csvTypeConvert(datasColumns,df_test,data_vocab)
 
     # 获取训练数据
     train_data = getDataSet(df_train,shuffleSize=trainSize,)