Generating data
Synthetic test data with deliberate outliers for system testing — and imputation when values are missing.
Synthetic test data
Sometimes you need data that doesn’t exist yet.
To evaluate a new CRM, a database or your application, we generate data to your specification: names, valid addresses, numeric attributes such as age, binary variables and more.
Deliberately placed outliers matter. A test only tells you something when you see how the system handles correct and incorrect input. We build both into the dataset.
Imputation
Imputation covers the statistical methods for filling in missing values. Not every method suits every kind of data — but it should be considered before any analysis or modelling.
We work closely with you here: only a real understanding of the data leads to values filled in correctly.
A simple example
| Name | Age | (other attributes) | Interested — raw | imputed |
|---|---|---|---|---|
| Wenger | 55 | … | Yes | Yes |
| Müller | 34 | … | — | Yes |
| Meier | 18 | … | No | No |
| Bauer | 27 | … | — | No |
Imputed values are shown in blue.
More services
Data cleansing
Cleanse, normalise, deduplicate and merge. For Swiss company addresses we can also match against the federal UID register.
Analysis
Customer bases, web analytics and quality analyses. Where are the hidden duplicates, which campaign actually worked.
Automations
Reports, recurring data preparation, transfers between systems. Also robotic process automation for forms and interfaces.
Surveys
Exploratory, descriptive or causal — from the hypothesis through the questions to an analysis that holds up.
Training & workshops
On every service, plus introductions to R, RStudio, R Shiny and Excel. Afterwards, you can do it yourself.
Custom requests
Your problem doesn’t fit any box? Tell us what you need — we’ll do our best to help.