Hi guys, this is my first post here!
I am a student from Italy and for my thesis I'm working with Knime.
I'm studying to become a laboratory technician and I'd like to do some analyzes about laboratory data.
Well, I've prepared an Excel CSV with 126000 rows, the columns are:
Exams - Exam result - Age - Gender - Exam Date - Exam Time
Now, I have to see the relationship there may be between the exam results and exam dates, and exam results and exam times but there is a problem: I can not put exam dates and exam results as a variables.
I've tried using String to Data/Time node but it didn't resolve my problem.
My data format is dd/MM/yy
So guys, how can I put exam dates and exam results on the x or y axis?
P.s. sorry for my bad English, for me it is unusual to speak of these things in another language.
If your exam dates are meant to be categorical varaibles, that is, each date independent from another and order between dates is not meaningful, then just use them as strings in a regression node. Similarly do the same with time. If you need the date as in result = alpha beta x date + noise where date is meant to be a continuous variable, then you will need to change it from datetime or string to a double or integer. ie,, result = alpha + beta x (date - start date) + noise to use one of the regression nodes.
I have laboratory data classified by type of examination (red blood cells ,
white blood cells , blood sugar and so on. ).
Each result of the laboratory has a precise time ( some are repeated ),
I would like to organize this data in order to see regressions and other things such as the application
of Bayes theorem.
The problem is that I do not know how to " prepare " the data for examination by KNIME .
I created using excel columns Exams ( type ) - Exam result - Age - Gender - Exam Dates - Exam Time .
They are all connected to each other of course .
I apologize if I keep still here but the main problem is to put the time on the x axis .
The file reader has time as a string, and the result as a number.
Then use the node "String To Date / Time" for the time .
After I connect "String To Date / Time" node with JfreeChartLine, I set the time on the X but when I run it appears this error: "Execute failed: You are attempting to add an observation for the time period Thu Jan 01 08:06:00 CET 1970 but the series already contains an observation for that time period. Duplicates are not permitted. Try using the addOrUpdate() method."
In the Csv files there are the dates of birth of patients ( prior to 1970, of course) but I skip the selected item . It could be the problem?
This is the list of results (Risultato) of a single laboratory test , in one day.
[IMG]http://i67.tinypic.com/2a774bc.jpg[/IMG]
I would put individual times on the X axis and see the pattern of results in a day.
After adding the String To Date / Time node add the JFreeChart Line and appears that message .
it maybe easier jus to create a table (the grid between the image and the omega sumbol on the top line of the tools) for the comment area and paste in the data