RPostgres | A DBI-compliant interface to PostgreSQL | Database library

glue_sql from glue package should work:

R was trying to connect to the wrong instance of postgres, after changing the port binding to something else (5432:5432 -> 8543:5432) I got connected to R.

How this happened I have no idea. It is my only running docker container, I dont have postgres on my host machine and connecting through the CL when bound to the standard port got me into the correct instance. Also I thought I would have gotten a port conflict from Docker if trying to use an in use port but that didn't happen either.

I think you need to:

1. identify which columns are missing,
2. alter table those new columns into existence, and then
3. upload the data, assuming all data in the first that are missing in the second are null-able.

To fix this, you first need to install "libpq-dev" as this is a prerequisite as per SO answer. Follow the steps below to apply this in a Google Cloud Notebook.

• Open a terminal in your JupyterLab and execute sudo apt-get install libpq-dev

• Open a R notebook, install Rpostgres using install.packages("RPostgres")

• Run library() and check if it was installed properly

Please take note that the machine I used for testing is created in AI Platform -> Notebooks -> R instance.

dbClearResult applies to the query object, not to the result of the fetch (which is a dataframe, hence the error message):

Maybe the driver you are using can't handle more than one statement per API call. Try doing the table and index creations in separate API calls:

As you can see from the article you provided, there's one requirement for you to communicate with your Cloud SQL instance locally and that is to authorize your public IP address in Cloud SQL.

This is similar and should be done as you will have to allow and authorize certain shinyapps.io IPs for it to communicate with your Cloud SQL instance.

To explain why, your client application's IP address or address range (on your case, shinyapps.io) must be configured as authorized networks when you're connecting using the public IP address of your Cloud SQL instance. This is a security design so unauthorized applications can't access your database easily.

The IPs that you need to allow can be seen from RStudio's support page. I will list it here:

Well, actually I still don't get a satisfying solution. What I exactly wanted to do is to drop containing NA rows in R environment without typing an SQL query, I think dbplyr doesn't support this function yet.

Then I wrote a little and simple code to make my wish come true;

You can use strings to pull the year subtract 1 and concatenate it back with the date and month. I am not sure if this will translate to sql though! This will also prevent leap years from messing up the days.