Drag and drop functionality

[cbkotura]

Hi David,

I had a few questions about the drag and drop plotting feature.

If I have data that is formated in this manner:

title 1
title 2
x y y y y
0 0 0 0 0
1 1 5 7 3
2 2 8 9 5

1. Is it possible for the drag and drop plot to label the y axis as y in this case if you have a single X and multiple Y values? Does this file need to be formatted differently so it can? I am hoping this is possible without having to change the way the text file is formatted, I am working with a large production database and it would not be possible to have the format changed most likely.
2. Is there a way to not have to click OK for every file that is dragged and dropped in, for instance if I drag 20 files at once is it possible to be plotted without clicking OK 20 times? That was the most common feedback given to me by my coworkers who I have asked to evaulate it as well.
3. If question 2 is possible, Is there any character that can be added above a column so that the column is ignored?

Thanks

[DPlotAdmin]

1. Is it possible for the drag and drop plot to label the y axis as y in this case if you have a single X and multiple Y values? Does this file need to be formatted differently so it can?

I'll change it such that if the Y column heading is the same for every column, that heading gets used as the Y axis label rather than the legend.

2. Is there a way to not have to click OK for every file that is dragged and dropped in, for instance if I drag 20 files at once is it possible to be plotted without clicking OK 20 times? That was the most common feedback given to me by my coworkers who I have asked to evaulate it as well.

Select the Open command on the File menu. Select file type D and click OK. On the following Open dialog, uncheck the "Pick Columns to Plot" box. That setting is persistent, so you won't see the "Specify Columns" dialog again unless you check that box.

3. If question 2 is possible, Is there any character that can be added above a column so that the column is ignored?

No, but that's something to think about for a future release. One thing you can do now is use a command line option to specify which columns to use. For example

dplot.exe /x1y3,4,5 filename.csv

will use the first column for X and skip column 2 (and anything after column 5). See "How do I specify options on the DPlot command line?" in the Help file for other command line options.

[cbkotura]

Hi David,

First off DPlot is a powerful and very useful engineering and scientific graphing tool with an attractive price. Just wanted to rehash some things I already posted before but also an example of the actually file format our database is setup as.

Select the Open command on the File menu. Select file type D and click OK. On the following Open dialog, uncheck the "Pick Columns to Plot" box. That setting is persistent, so you won't see the "Specify Columns" dialog again unless you check that box.

I tried that and it worked fine. Until I reopened DPLOT. After that all points in the x column (column 1) were plotted as line, and all points in the y column were plotted as a line, and so forth. The only way I found that you can fix it is to check "specify colums" again, and then unselect it again each time.

But some issues/suggestions our company has are the following:

Unfriendly to drag-and-drop 2D multi-column (XYYYYY...) data.
Can't automatically assign Y-axis title in this data format.
Can't arbitrarily ignore which Y column to plot without manually selecting it when opening this data format.

Please implement this functionality of easily reading a 2D multi-column data format (XYYYYY....) used mostly in the mass-production industry. This format has been widely accept in many database systems as well. The standard DPlot file (*.GRF) is not friendly for data processing and database recording purposes. An example of the data file is given here:

----------------------------------------------------------
'DPlot should automatically ignore comment or note without
'any manual settings required.
'Wafer ID: WKIPOWXC091
'Device ID: K093UI22
'Tested Date: 11/09/2009
'Tested Time: 15:03
'Testing station: PD7
'Operator: John C.
'Pass or Fail: Pass

Diode Array IV '1st Title line
Device-K093UI22 '2nd Title line
Bias (Volt) 'X-axis title (no need for a 3rd title line)
Current (mA) 'Y-axis title (must have a way to control y axis title)
XYY..Y 'Plot only 2, 3, and 6 Y-columns

Volt Ch1 Ch2 Ch3 Ch4 Ch5 'Y column headings used as curve legend
0.12 10 11 13 15 18
0.28 7 6 44 89 23
0.33 4 5 22 32 25
0.45 8 12 77 68 30

'Other DPlot settings followed
.....

Thanks again, and DPLOT really is one of the best graphing tools we have ever come across.

[DPlotAdmin]

I tried that and it worked fine. Until I reopened DPLOT. After that all points in the x column (column 1) were plotted as line, and all points in the y column were plotted as a line, and so forth. The only way I found that you can fix it is to check "specify colums" again, and then unselect it again each time.

If this is really what you get then DPlot is somehow misinterpreting your file. You're welcome to send me a file that you're having a problem with. When I copy your example and save to a text file, then open it with DPlot, I get the expected results: X starting at 0.12, ending at 0.45, and 5 curves with legends Ch1, Ch2, etc.

XYY..Y 'Plot only 2, 3, and 6 Y-columns

I can't see the value in this. You're writing this line in your file hoping to tell DPlot which columns to plot. But if you don't want to plot the other columns, why write them to the file in the first place? I can see how you might want to plot some of the columns under some circumstances and other columns under other circumstances. But you'd have to edit the file to make that change, rather than clicking a few buttons in DPlot with "Pick Columns to Plot". I don't mind editing files myself, but I can't think of a reason why anybody would think this would be easier.

[cbkotura]

I can't see the value in this. You're writing this line in your file hoping to tell DPlot which columns to plot. But if you don't want to plot the other columns, why write them to the file in the first place? I can see how you might want to plot some of the columns under some circumstances and other columns under other circumstances. But you'd have to edit the file to make that change, rather than clicking a few buttons in DPlot with "Pick Columns to Plot". I don't mind editing files myself, but I can't think of a reason why anybody would think this would be easier.

'DPlot should automatically ignore comment or note without
'any manual settings required.
'Wafer ID: WKIPOWXC091
'Device ID: K093UI22
'Tested Date: 11/09/2009
'Tested Time: 15:03
'Testing station: PD7
'Operator: John C.
'Pass or Fail: Pass

Diode Array IV '1st Title line
Device-K093UI22 '2nd Title line
Bias (Volt) 'X-axis title (no need for a 3rd title line)
Current (mA) 'Y-axis title (must have a way to control y axis title)
XYY..Y 'Plot only 2, 3, and 6 Y-columns

Volt Ch1 Ch2 Ch3 Ch4 Ch5 'Y column headings used as curve legend
0.12 10 11 13 15 18
0.28 7 6 44 89 23
0.33 4 5 22 32 25
0.45 8 12 77 68 30

'Other DPlot settings followed

In our case our we have multiple diode arrays (ch1, ch2, etc..) on a single chip. All these arrays are tested by an operator at one time and the data txt file is automatically generated with those headerlines shown and stored into a database we could easily change the program to add more headerlines for dplot commands but we still would want the data on all the different channels in one file per chip. In this case Ch1, Ch2, and Ch6, are the arrays that we would like to show the data for to a customer always since these are the production channels. Ch3 and Ch4 are R&D channels and we do not wish to show a customer the results on those channels. Clicking which column you want to plot is fine if a customer orders a small number of chips. But in our case a customer is ordering thousands at a time and click which column would be time consuming so we were just trying to find a solution to make this step more efficient for us. For the R&D channels clicking choices is not as bad since those are not look at on a regular basis and when they are analyzed probably less than 20 are anazlyed.

[DPlotAdmin]

Thanks for the details.

I think the best solution here will be to write a file import plugin. This will avoid fouling things up with other users. For example you consider ' to be an indication of a comment that should be ignored. But there's nothing magic about that character; it is perfectly valid in a column heading, for example. Similarly, there's nothing magic about the character sequence XYY.

The only problem with relying on an import plugin is you need a character string within the file that positively identifies it as being of the correct type. Otherwise drag-and-drop and "Open with..." will not work.