Date: 2017-11-17 15:15 |
Priority: more information 5 |
Submitted By:
Sarthak Pati (sarthakcbica)
|
Assigned To: more information
Saima Rathore (saimarathore) |
Component: Applications: Survival Prediction Index | Version: None |
Severity: None | Resolution: Awaiting Response |
Operating System: All |
Summary: more information [Christos] Feedback about Survival |
Detailed description |
- testing Survival module: specified model directory doesn't
exist. The model doesn't seem to be there in the distribution or
sample data. NOT WORKING
- More on Survival module: Using existing model, specifying
the folder. Specifying test data. Running—crashes the entire
s/w.
| |
|
Response
Message |
Date: 2017-12-14 19:28
Sender: Sarthak Pati
New training works fine.
|
Date: 2017-12-14 19:08
Sender: Sarthak Pati
Testing on Existing Model:
- No longer crashes (Beta.g)
- What is the significance of SPI? Basic information would be
nice
- The information that was used to generate the result needs to
be restructured: first should be model directory, then the input
directory and finally the output directory (the way the pop-up
dialog is structured); also, this information should be saved in
the output directory (especially considering that we are asking a
user for an output directory and not a single file)
- In the output directory, there is a single file 'results.csv',
which has 2 columns: subjectID and Result; but there are 4 fields
getting populated from the next line onwards, for example, the
first line is: AAAA,2.952639,-0.808882,0.191118. This is
confusing and needs to be clarified with different headers
- It is mentioned that the SPI is saved in results.csv; but there
is nothing there referencing this. If someone runs Survival and
generates results, they will not be able to correlate the SPI
information (and how is it calculated for X number of
subjects)
| |
|
Attached Files:
Name |
Download | No Files Currently Attached | |
|
Changes:
Field |
Old Value |
Date |
By |
status_id | Open | 2018-04-16 18:10 | sarthakcbica |
New Message | | 2017-12-14 19:28 | sarthakcbica |
Message | Testing on Existing Model:
- No longer crashes (Beta.g)
- What is the significance of SPI? Basic information would be
nice
- The information that was used to generate the result needs to
be restructured: first should be model directory, then the input
directory and finally the output directory (the way the pop-up
dialog is structured)
- In the output directory, there is a single file 'results.csv',
which has 2 columns: subjectID and Result; but there are 4 fields
getting populated from the next line onwards, for example, the
first line is: AAAA,2.952639,-0.808882,0.191118. This is
confusing and needs to be clarified with different headers
- It is mentioned that the SPI is saved in results.csv; but there
is nothing there referencing this. If someone runs Survival and
generates results, they will not be able to correlate the SPI
information (and how is it calculated for X number of
subjects)
| 2017-12-14 19:15 | sarthakcbica |
New Message | | 2017-12-14 19:08 | sarthakcbica | |
|