Is there a better way to batch copy multiple columns ？

[#1]

Hi everyone! I noticed that when using "Paste with column names" to copy multiple columns, selecting multiple columns in the left column window doesn't work. You actually have to select these columns directly in the data window to use it normally. But this makes it really inconvenient to operate, especially when there are a whole bunch of scattered columns. Is there a better way to copy multiple columns in batches? Thanks!

I'm using version 18.2.2.

[#2]

I'm not exactly sure what you are trying to do here.

If you select columns from the Column panel it will also select them from the data table

If you wish to copy the data to clipboard, you also have to select the rows you wish to copy and you have to do it after selecting the columns from the left. Otherwise, it will just copy the column names as that is the active selection.

[#3]

I think I understand the working mechanism you are explaining.

After selecting the column names in the column panel, the columns in the data panel will be in a light blue highlighted state. At this time, I need to click on the blank area above the row number list in the data panel to make the selected columns turn into a dark blue highlighted state. Only then can it be said that the columns and the data inside them have been selected.

Only when source and target columns all in dark blue state can these two commands be used to copy and paste data.

Thank you very much for your timely guidance!

[#4]

You can click anywhere in the data table.

Here you will have just columns selected and no data AND selection is active on column names

You can click anywhere on the table, but depending on where you click different selections will happen. Just make sure that the columns stay selected when you click on the table and figure out which option is best for you (blank space somewhere, row listing or holding ctrl and clicking on the table cells).

If you click on blank space, the selection will change from column names to column headers

I think you can copy at this stage but you can also select the rows you are interested in (ctrl+A to select all rows).

[#5]

Jmp has about the highest complexity I've seen for basic operations like copying columns.

Fortunately, there are masters like you to provide detailed guidance, thank you very much!

[#6]

It can feel complex sometimes but, in this case, there are two separate selections you are doing: column names and column headers. Most likely there could be some improvements which could be done to make it more clear... not sure if this will change in JMP19 for example.

I will tag @julian so she sees that there can be some confusion when user is selecting columns from columns panel. Maybe for example "Copy with column names" could be expanded to always take into account the data.

[#7]

---

@jthi wrote:

You can click anywhere on the table, but depending on where you click different selections will happen. Just make sure that the columns stay selected when you click on the table and figure out which option is best for you (blank space somewhere, row listing or holding ctrl and clicking on the table cells).

---

This is so cool in JMP!
Very unique - and powerful: the column selection is 'conserved' when the user selects specific rows.
And vice versa: when a user selects a column, this doesn't change or reset the row selection.
Once you get used to it you don't want to miss it! Guaranteed! *)

In Excel, when you select a row, the column selection disappears.
The same happens when you select a column: the row selection is lost.

Not convinced? Just watch some of the amazing lectures by @brady_brady  ( https://www.google.com/search?q=data+doctor++site%3Acommunity.jmp.com  ) -  When data curation goes beyond copying columns, JMP almost looks like magic! Some of this just works due to the powerful way in which rows and columns can be selected in JMP.

[#8]

---

@jthi wrote:

Maybe for example "Copy with column names" could be expanded to always take into account the data.

---

At the moment, "Copy Columns with column names" follows the idea

- just the selected columns get copied

- just the selected rows get copied

It works like "Copy Columns" via column header/right click:

 Copy Columns - selected rows 

One can argue that many users don't need the current functionality.

On the other hand, I am sure, there are some (long term?) JMP users who use and love the current way it works :

If there are plans to change the way how it works, there should be an option in the preferences which allows the existing users to use the existing functionality.

[#9]

Sorry, there is one important detail in my previous post that is wrong:
Actually, there is a significant difference between "Copy Columns with column names" and "Copy Columns" via the column header right click menu:

"Copy Columns with column names" 

1. just the selected columns get copied
2. just the selected rows get copied *)

*) edit: 

3. if no row is selected, all rows get copied. 
(already mentioned by Jarmo in a previous post: "I think you can copy at this stage")

so, "Copy Columns with column names" already behaves like suggested here:

@jthi wrote:

Maybe for example "Copy with column names" could be expanded to always take into account the data.

---

On the other hand:  "Copy Columns" via Column Header / Right Click

3. if no row is selected, no entry gets copied *)

 *) Not useful?
On the contrary! This function can be used to create a script in the clipboard that generates the column structure (including all properties). Extremely useful for inserting the script into a workflow afterwards — without the ballast of column entries.

[#10]

Another difference is speed.

The video below shows the 3 cases:

- "Copy Columns with column names" / "Paste Columns with column names" (just the entries, no column properties)

- "Copy Columns" / " Paste Columns": no row selection  (just the column structure is created)

- "Copy Columns" / " Paste Columns": all rows selected -> column structure + entries

The 3rd approach is very comfortable:
It can be used to create precise copies of the original columns - including all column properties AND the values.

But the user has to pay for the comfort - with time.
"Paste columns" takes time - like in the video!
For larger tables, it can take minutes or hours to create the new columns. 

If you want to copy and paste columns, including all their properties and entries,

... and the data table is large,
You can save hours by doing it in two steps:
Either: 

- copy/paste the structure via Copy Columns & Paste Columns (with no rows selected)

- copy/ paste the content via CTRL+C & CTRL+V

Or vice versa:

- "Copy Columns with column names" & "Paste Columns with column names" (just the entries, no column properties)

- add the column properties via Column Header / right Click / "Copy Columns Properties" & "Paste Columns Properties"

@NoFinalAnswer wrote:

Jmp has about the highest complexity I've seen for basic operations like copying columns.

---

agree ...

There are many different ways to complete tasks in JMP. And for a specific purpose, one approach can be much better than another.
For a slightly different purpose, the opposite may be true.
It's not always easy to know or remember which approach is the right one, and sometimes the user ends up with a frozen program as a result.