A complete database research workflow (Matlab & RStudio)
Last updated
Was this helpful?
Last updated
Was this helpful?
Nuvolos supports a query-analyze-insert cycle to help researchers query, analyse and store data safely in a simple, integrated workflow. The workflow is demonstrated in two languages, but can be extended to any languages supported by Nuvolos.
A standard skeleton scientific workflow in Matlab can be broken down into three main steps:
Query research-relevant data
Analyse, transform or otherwise manipulate data
Store results
For the mock example, we are going to work with the Fama-French factor set that is available for our Demo user, we will be focusing on the 5-factor table. The NORTH_AMERICA_5_FACTORS table has been .
After opening the Matlab application, the following bit of code will return the entire database table as a Matlab table-type object. The query we are executing is a merge of a monthly stock series table (for Apple monthly stock prices) and the Fama-French factor table as follows.
The SQL query:
The code that executes the query, the above string is saved in query_string.
The previous step resulted in dataset_factor containing a Matlab Table object that holds the data. The fitlm method fits a linear regression on the table with an R-style formula. We then write back the fitted values to the Table as a column.
As a final step, we write back the results using the data upload command for Matlab:
As an end result, we can then find the table on the Nuvolos UI:
A standard skeleton scientific workflow in RStudio can be broken down into three main steps:
Query research-relevant data
Analyse, transform or otherwise manipulate data
Store results
It is possible to create more complext workflows, but it will usually consist of three-step modules such as above.
After opening the RStudio application, the following bit of code will return the entire database table as an R data frame object. The query we are executing is a merge of a monthly stock series table (for Apple monthly stock prices) and the Fama-French factor table as follows.
The SQL query:
The code that executes the query, the above string is saved in query_string.
The previous step resulted in dataset_factor containing an R data.frame object that holds the data. The lm method fits a linear regression on the data frame. We put the fitted values to the data frame.
As a final step, we write back the results using the data upload command for R:
For the mock example, we are going to work with the Fama-French factor set that is available for our Demo user, we will be focusing on the 5-factor table. The NORTH_AMERICA_5_FACTORS .
