Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

It is possible to train multiple file at once using same learning algorithm?

Hi,

Let say I have 10 days excel data in separate file, everything inside the file is same such as column name and their data type. Only each value(rows) and number of rows may different.
To save time, it is possible to run all this file with 1 click?

Thanks.

Find more posts tagged with

AI Studio

Excel

Time Series

Loops + Branches

Tutorials + Training

Auto Model

Accepted answers

MarlaBot

Hi @muhd_ikram - this is MarlaBot. I found these great videos on our RapidMiner Academy that you may find helpful:

Instructional Video: 'Replace Missing' Values in Time Series (Viewing time: ~4m)

Instructional Video: Time Series Forecasting and Feature Generation (Viewing time: ~6m)

Instructional Video: Time Series Forecast and Validation with ARIMA (Viewing time: ~5m)

Instructional Video: Using windowing on Time Series data (Viewing time: ~5m)

Instructional Video: Time Series calculations with 'Process Windows' (Viewing time: ~9m)

Please LIKE my comment if it helps! 👇

MarlaBot

Papad

Hi @muhd_ikram
You can use "Loop Files" operator. You have to include all files you want from parameters and then connect this operator with the rest of the flow you've made. What is going to happen is that every time your flow will run for each of your files and give the same number of results.
Hope it helps,
Jim.

varunm1

Hello @muhd_ikram

Based on your post, if all the files are in same format with same attribute names (column names), you use loop files with a read operator inside loop files (read csv or read xls based on your file format) then connect the loop files operator to "append" operator and then you can create the rest of modeling process.

If the files are named with a naming convention its fine but if they are not I recommend you name them file_01, file_02,.. file_10, this will help loop files read in sequence.

Note: append operator work only if your attribute names are same.

Telcontar120

You don't even need the append, if you want to run the model on each file separately. If they really are all the same, then you can just do everything inside the Loop Files and output the collection of model results.

All comments

MarlaBot

Hi @muhd_ikram - this is MarlaBot. I found these great videos on our RapidMiner Academy that you may find helpful:

Instructional Video: 'Replace Missing' Values in Time Series (Viewing time: ~4m)

Instructional Video: Time Series Forecasting and Feature Generation (Viewing time: ~6m)

Instructional Video: Time Series Forecast and Validation with ARIMA (Viewing time: ~5m)

Instructional Video: Using windowing on Time Series data (Viewing time: ~5m)

Instructional Video: Time Series calculations with 'Process Windows' (Viewing time: ~9m)

Please LIKE my comment if it helps! 👇

MarlaBot

Thanks everyone, I tried and it works. I just wonder, since it already can read and produce result separately, how we can combine all these results into 1? such as all 10 different prediction results in 1 visualization for comparison.

varunm1

@muhd_ikram

Can you provide your XML code from XML window (View --> Show Panel --> XML)? Based on how you created the process, we can inform you how to combine all the results. You can post here based on the following screenshot.

Thanks