Hi,

I recently got an interesting question from an SPSS user. They wanted to let their own customers access a dataset through the web and analyze it themselves, generating percentage crosstables – while SPSS would run the actual analyses in the background. (Sort of like pivoting in Excel).

Is this possible?

/Gunilla Rudander

Hi @Gunilla Rudander. It should be possible, provided you can make the dataset locally available to Statistics. That can be done either by uploading the dataset to a file system that Statistics has access to, or by putting the dataset in a database and having Statistics access the data that way.

What you need is to drive a copy of Statistics using either Python or the R language. Then when your users upload a dataset and ask for an analysis, you would have the Python or R code submit syntax to open the dataset, run the analysis, and export the results to a format that you can deliver on the web such as HTML. You could either do this using OMS syntax to redirect the output or by issuing OUTPUT EXPORT syntax directly from R or Python.

HTH,

Hi,

I recently got an interesting question from an SPSS user. They wanted to let their own customers access a dataset through the web and analyze it themselves, generating percentage crosstables – while SPSS would run the actual analyses in the background. (Sort of like pivoting in Excel).

Is this possible?

/Gunilla Rudander

Thank you for your reply, Curtis. This is a good starting point, but this company would like to be able to offer their customers the possibility to access a webpage and download statistics in the form of multi-level cross tables, for example salaries splitted by years in the labor market, age, or industry. So I believe the SPSS data file needs to be stored in some cloud-based location, rather than being downloaded locally by their customers as you suggested. /Greetings Gunilla Rudander

Hi @Gunilla Rudander. It should be possible, provided you can make the dataset locally available to Statistics. That can be done either by uploading the dataset to a file system that Statistics has access to, or by putting the dataset in a database and having Statistics access the data that way.

What you need is to drive a copy of Statistics using either Python or the R language. Then when your users upload a dataset and ask for an analysis, you would have the Python or R code submit syntax to open the dataset, run the analysis, and export the results to a format that you can deliver on the web such as HTML. You could either do this using OMS syntax to redirect the output or by issuing OUTPUT EXPORT syntax directly from R or Python.

HTH,

Hi,

I recently got an interesting question from an SPSS user. They wanted to let their own customers access a dataset through the web and analyze it themselves, generating percentage crosstables – while SPSS would run the actual analyses in the background. (Sort of like pivoting in Excel).

Is this possible?

/Gunilla Rudander

Hi @Gunilla Rudander,

Apologies if there was some confusion. By 'locally', I meant local to the file system that SPSS Statistics has access to. Users can upload CSV files or some other format, SPSS Statistics can read and process those files, then export the output to some format that users can consume. At no time would datasets need to be downloaded to the user's local machine.

Best,  

Thank you for your reply, Curtis. This is a good starting point, but this company would like to be able to offer their customers the possibility to access a webpage and download statistics in the form of multi-level cross tables, for example salaries splitted by years in the labor market, age, or industry. So I believe the SPSS data file needs to be stored in some cloud-based location, rather than being downloaded locally by their customers as you suggested. /Greetings Gunilla Rudander

Hi @Gunilla Rudander. It should be possible, provided you can make the dataset locally available to Statistics. That can be done either by uploading the dataset to a file system that Statistics has access to, or by putting the dataset in a database and having Statistics access the data that way.

What you need is to drive a copy of Statistics using either Python or the R language. Then when your users upload a dataset and ask for an analysis, you would have the Python or R code submit syntax to open the dataset, run the analysis, and export the results to a format that you can deliver on the web such as HTML. You could either do this using OMS syntax to redirect the output or by issuing OUTPUT EXPORT syntax directly from R or Python.

HTH,

Hi,

I recently got an interesting question from an SPSS user. They wanted to let their own customers access a dataset through the web and analyze it themselves, generating percentage crosstables – while SPSS would run the actual analyses in the background. (Sort of like pivoting in Excel).

Is this possible?

/Gunilla Rudander