Skip to content

Throw Exception"genTF unsupported label data type: DOUBLE_TYPE" #1166

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Echo9573 opened this issue Nov 10, 2019 · 4 comments · Fixed by #1199
Closed

Throw Exception"genTF unsupported label data type: DOUBLE_TYPE" #1166

Echo9573 opened this issue Nov 10, 2019 · 4 comments · Fixed by #1199
Assignees
Labels
bug Something isn't working DataScience The issue about the application in data science DiDi The issue publisher is from Didi

Comments

@Echo9573
Copy link
Collaborator

Echo9573 commented Nov 10, 2019

Description
Throw Exception "converting driver.Value type float64 ("-0.034224987") to a int64 " when select data from predictions in notebook.

Reproduction Steps
image

%%sqlflow
SELECT *
FROM intelligence_da_dev.titanic_train
TRAIN DNNClassifier
WITH
  model.n_classes = 2,
  model.hidden_units = [10, 40, 10],
  train.epoch = 200,
  train.batch_size = 64
COLUMN age, fare, pclass_3, pclass_1, pclass_2, sex_male, sex_female, sibsp_1, sibsp_0, sibsp_others_type, parch_0, parch_1, parch_2, parch_others_type, embarked_s, embarked_c, embarked_others_type, embarked_unknown_type
LABEL survived
INTO intelligence_da_dev.my_dnn_model;

--------------------------------Exception------------------------------
_Rendezvous: <_Rendezvous of RPC that terminated with:
status = StatusCode.UNKNOWN
details = "genTF unsupported label data type: DOUBLE_TYPE"
debug_error_string = "{"created":"@1573309027.737003066","description":"Error received from peer ipv4:10.82.128.7:8005","file":"src/core/lib/surface/call.cc","file_line":1055,"grpc_message":"genTF unsupported label data type: DOUBLE_TYPE","grpc_status":2}"

@Echo9573
Copy link
Collaborator Author

This question is similar to #770

@Echo9573
Copy link
Collaborator Author

Hi, @Yancey1989, When I manually convert the double type to a float, another exception is also thrown, as followings shows. Maybe it is necessary for SQLFlow to support common database types.

--------------------------------Exception------------------------------
_Rendezvous: <_Rendezvous of RPC that terminated with:
status = StatusCode.UNKNOWN
details = "genTF unsupported label data type: BIGINT_TYPE"
debug_error_string = "{"created":"@1573389345.693687540","description":"Error received from peer ipv4:10.82.128.7:8005","file":"src/core/lib/surface/call.cc","file_line":1055,"grpc_message":"genTF unsupported label data type: BIGINT_TYPE","grpc_status":2}"

@Yancey0623 Yancey0623 added the bug Something isn't working label Nov 10, 2019
@Echo9573
Copy link
Collaborator Author

The hive data type is as follows:
Tinyint, smallint, int, bigint, boolean, float, double, string, binary, timestamp, decimal, date, varchar, char, array, map, struct.
Common data types may have at least tinyint, smallint, int, bigint, float, double, decimal, varchar, char

@Yancey0623
Copy link
Collaborator

Yancey0623 commented Nov 10, 2019

Hi @Echo9573 , thanks for your report, I will reproduce your case.


The hive data type is as follows:
Tinyint, smallint, int, bigint, boolean, float, double, string, binary, timestamp, decimal, date, varchar, char, array, map, struct.
Common data types may have at least tinyint, smallint, int, bigint, float, double, decimal, varchar, char

It seems that array/map/struct is more complex, can you show some cases that how to use them in Tensorflow/XGBoost or other toolkit ?

@Echo9573 Echo9573 added DataScience The issue about the application in data science DiDi The issue publisher is from Didi labels Dec 2, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working DataScience The issue about the application in data science DiDi The issue publisher is from Didi
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants