Automate data processing and deliverables from live Nebu data
Datasmoothie’s Tally, the API for survey data, helps bring online surveys into your data pipeline
Datasmoothie recently launched Tally, a RESTful API for survey data. One of the integrations we implemented very early on is for the impressive survey platform Nebu.
Nebu is a powerful suite of survey data tools, with almost thirty years of experience in building and delivering software solutions for fieldwork and market(ing) research companies.
Connecting to Nebu
Using Tally to connect to your Nebu data couldn’t be easier. Using the Python client to Tally (pip install datasmoothie-tally-client) we write:
import tally
dataset = tally.DataSet(api_key=tally_api_key)
dataset.use_nebu(live_link_to_nebu_sav)Once the connection has been made, the user can use Tally to perform common data processing tasks (more on that here) and to produce deliverables, such as Excel Tables and PowerPoint decks.
Building deliverables
PowerPoint
Tally can generate branded PowerPoint decks with your results with one line of code. The user can send a branded PowerPoint template to Tally and it will use that to generate PowerPoint files with the variables requested.
questions = ['Q2', 'Q3', 'Q4.1', 'Q4.2', 'Q5b']
crosstabs = ['@', 'Gender', 'Q1']dataset.build_powerpoint(filename='result.pptx',
powerpoint_template='branded-template.pptx',
x=questions,
y=crosstabs)
Filtering is also supported, so looping over categories of a specific variable (e.g. countries, stores) and creating one PowerPoint for each is easy.
The PowerPoint that is generated has one chart for each question and one chart for each question, crossed with the requested variables (@ means total). In this example, we used a template with the Datasmoothie colours and logos, but users can upload any template they want to brand to their agency or their client.
Excel tables
Building Excel tables is just as easy. We send the same variables for questions and cross-tabs and save the output as an Excel file. The Excel table generation supports sig-testing, filters and various different visual options.
dataset.build_excel(
filename='excel_output/nebu.xlsx',
x=questions,
y=crosstabs)
Bringing it together in a pipeline
The above shows how both the data processing and the generation of tables and PowerPoint files can be automated and integrated into any data pipeline. This can be particularly useful for trackers, where emailing results in tables and decks can be done at a regular frequency, without any manual input.
