on 06-01-2015 10:05 AM
Hello,
I am new to PAL. I am using APLWRAPPER_FORECAST function to predict the amount for future years. My input dataset contain ID, Year and Amount. Currently I am predicting amount for each Id separately and then combining the result. Could I directly give ID and Year as a key in input and predicted result should also contain the same.
Thanks.
Hi Mayur,
Can you please provide us sample data? It looks like you have only one row in your dataset.
Regards
Angad
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello Angad,
I have enough data to train the model and I am able to produce results with single ID.
My dataset is like below
ID(Post_Code)---------Year------------Amount
ID is not continues and the key is ID+Year.
I have 5 IDs and amount for each ID from 2003 to 2026. I am collecting data for a particular ID in input table and generating result as Year and Amount. Hence the result dataset is only for that particular ID.
Instead of collecting data for each ID I want to give whole table as input and result should also contain ID,Year and predicted amount for future years. But I am not able to include the ID(post_code) in my output data set.
Below are the table types of Input and Output and FORECAST function which I am currently using:
create type post_code_data_input_t as table(
"post_code" integer, (ID)
"year" integer,
"amount" double
);
create type post_code_data_output_t as table(
"year" integer,
"amount" double,
"kts_1" double
);
call APLWRAPPER_FORECAST(FUNC_HEADER, FORECAST_CONFIG, VARIABLE_DESC, VARIABLE_ROLES, POST_CODE_DATA_INPUT_TBL, POST_CODE_DATA_OUTPUT_TBL, OPERATION_LOG, SUMMARY, INDICATORS) with overview;
Inpute Dataset is:
post_code Year Amount
2540 2003 8000.123
2540 2026 7000.223
.
.
.
.
2540 2016 9000.324
And similarly for other 4 post codes.
Regards,
Mayur
User | Count |
---|---|
86 | |
10 | |
10 | |
9 | |
6 | |
6 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.