Comments on: SQL Server Machine Learning Services – 3 Ways to Run ML Workloads

By: SQL On-Demand (Serverless Query) in Azure Synapse Analytics | @Subaru

SQL On-Demand (Serverless Query) in Azure Synapse Analytics | @Subaru — Sat, 23 May 2020 22:42:48 +0000

[…] due to architectural reasons. For instance, T-SQL in Synapse Analytics supports PREDICT (see here), with which you can infer values with a trained model by machine learning. But, this might not be […]

LikeLike

By: Yulin Wu

Yulin Wu — Tue, 13 Nov 2018 15:29:30 +0000

Hi Tsuyoshi,

When installing SQL 2017 instance without the ”Advanced Analytics Extensions”, I’m not able to use sp_execute_external_script from SSMS. That is expected.

However, when I’m trying to use RevoScaleR from my RStudio locally, I can have the remote compute context established as below. My question is without the “Advanced Analytics Extensions” installed remotely, is RevoScaleR still using the remote SQL Server as the compute context? I thought it won’t have the R runtime required. Is there a simple example I can test which context is used?

> rxGetComputeContext()
RevoScaleR Microsoft SQL Server In-Database Object
————————————————–
connectionString : “Driver=SQL Server;Server=XXXXX;Database=test_Dream;Trusted_Connection=True;”
numTasks : 1
executeAsUser : “”
executionTimeoutSeconds : 0
shareDir : “C:\\AllShare\\XXX”
wait : TRUE
consoleOutput : FALSE
autoCleanup : TRUE
nodes : NULL
packagesToLoad : NULL
description : “sqlserver”
version : “1.0-2”

LikeLike

By: Run deep learning workloads on SQL Server Machine Learning Services (rxNeuralNet) – Tsmatz

Mon, 19 Mar 2018 10:23:12 +0000

[…] my previous post I explained several approaches and what you need to know about SQL Server in-database machine […]

LikeLike