gs_new() and gs_ws_new() too slow

[ankhnesmerira]

As of 2022-02-03, the googlesheets package has been archived on CRAN. I'm closing all issues and pull requests prior to archiving this GitHub repo.

Learn more at:

https://github.com/jennybc/googlesheets#readme

[jennybc]

Wow! Working with Google Sheets via API is not performant but what you're describing is very unusual.

Here's what I just got with gs_new():

> system.time(iris_ss <- gs_new("iris", input = head(iris, 3), trim = TRUE))
Sheet "iris" created in Google Drive.
Range affected by the update: "R1C1:R4C5"
Worksheet "Sheet1" successfully updated with 20 new value(s).
Accessing worksheet titled 'Sheet1'.
Sheet successfully identified: "iris"
Accessing worksheet titled 'Sheet1'.
Worksheet "Sheet1" dimensions changed to 4 x 5.
Worksheet dimensions: 4 x 5.
   user  system elapsed 
  0.919   0.034  10.670 

I have never written a data.table this way but also don't see why that would affect things. If you write a plain vanilla data frame, do you see a difference?

[ankhnesmerira]

Hi, thanks for your response. I've tried with plain data.frame and it's the same issue. I've put the question online as well and it seems there are other people who are experiencing the same slow performance.

thanks

[tomwwagstaff-product]

I am experiencing the same issue with gs_ws_new() - writing datasets of modest size (a few columns, a few hundred rows) is taking perhaps 10 minutes...

However, this only seems to be a problem when I'm uploading dplyr tibbles - I can upload a vanilla dataframe very snappily.

P.S. love the googlesheets package - it's making my life so much easier - thank you :-)

[jennybc]

Hmm ... interesting re: tibble vs. data frame. In any case, the reboot googlesheets4 is well underway, with reading public Sheets already available for use. Writing will come soon!

What I'm saying: I will not be troubleshooting this any further here and, I hope, it will simply not be a problem in googlesheets4, with the new API.

[mlamias]

I too am having this problem, but I'm using a plain-old non-dplyr'ed data frame that is 3038 x 26 (not exactly massive. I'm hoping there's a resolution soon! Love the package!

[hiteshahuja29]

Hey Jenny, can we please have an update on the new googlesheets package? Waiting eagerly for this. The googlesheets package has a very slow gs_edit function.

[jennybc]

I'm back on this google stuff now. I have to get gargle onto CRAN first (the package providing auth), then googlesheets4. I think it's likely that googlesheets4 will go to CRAN first as read only (sort of it's current form, really), just to get the ball rolling sooner, but YES writing into Sheets from googlesheets4 is a very high priority.

[luispitta]

I am experiencing the same issue with gs_ws_new() - writing datasets of modest size (a few columns, a few hundred rows) is taking perhaps 10 minutes...

However, this only seems to be a problem when I'm uploading dplyr tibbles - I can upload a vanilla dataframe very snappily.

P.S. love the googlesheets package - it's making my life so much easier - thank you :-) Hi Tom, what's a Vanilla Data Frame? Can you explain to me? I try with a normal df and its too slow. My DF has 35 columns and 179 rows

[tomwwagstaff-product]

Hi @luispitta, Yeah I just meant a normal dataframe, sorry that's not working for you

[luispitta]

Wow! Working with Google Sheets via API is not performant but what you're describing is very unusual.

Here's what I just got with gs_new():

> system.time(iris_ss <- gs_new("iris", input = head(iris, 3), trim = TRUE))
Sheet "iris" created in Google Drive.
Range affected by the update: "R1C1:R4C5"
Worksheet "Sheet1" successfully updated with 20 new value(s).
Accessing worksheet titled 'Sheet1'.
Sheet successfully identified: "iris"
Accessing worksheet titled 'Sheet1'.
Worksheet "Sheet1" dimensions changed to 4 x 5.
Worksheet dimensions: 4 x 5.
   user  system elapsed 
  0.919   0.034  10.670 

I have never written a data.table this way but also don't see why that would affect things. If you write a plain vanilla data frame, do you see a difference?

Hi Jenny, im trying to create a google sheet with my df, but its really slow and i don't know the reason, i have a tibble with 35 columns and 178 rows.

[luispitta]

Hi Guys, i resolve the problem by reducing the columns in my data frame, i also use the parameters: col_extent = ncol(data_frame), row_extent = nrow(data_frame) Hope it can help you to.