Forcing column headers on import, using regression [message #184] |
Tue, 26 July 2016 03:13 ![Go to next message Go to next message](/forum/theme/myThemeWithHFCSS/images/down.png) |
skarrila
Messages: 1 Registered: July 2016
|
Junior Member |
|
|
I use this package for exploratory data visualization, not for cheminformatics.
To de-pivot I need to have the first row of my cut/paste Excel data as headers. I managed to force this by making an extra column with string in first row and numbers after that; otherwise I got generic column names. Is there some built-in feature for the purpose of forcing a row to become column headers? If not, how can I effectively cut/paste a row to become headers?
The package is excellent in what it does, but I can't access the least squares regression with my data. If you simply have a column "y" as targeted output, and columns "x1",..., "xn" as model inputs, how is linear regression executed to fit a predictive model? It seems that the regression function is currently "hidden" so it can only be used with computed chemical characteristics...
Thank you and best regards.
Seppo
|
|
|
Re: Forcing column headers on import, using regression [message #185 is a reply to message #184] |
Tue, 09 August 2016 13:36 ![Go to previous message Go to previous message](/forum/theme/myThemeWithHFCSS/images/up.png) |
thomas
Messages: 693 Registered: June 2014
|
Senior Member |
|
|
Sorry for the delay. I see the point and added two menu options to enforce pasting with and without a header row. The 'Paste' macro task now also has these options.
To your second point: I had built in a PLS regression only temporarily and just for the purpose of calculating atom type based contributions for polar and non-polar surface prediction, but I didn't invest the time to build an UI, test, and document it to open it for general model building. If there is sufficient interest, I may add it after the upcoming DataWarrior update.
Kind regards,
Thomas
|
|
|