Product functionality

This page provides a detailed overview of our software suite, including the features of our platform and the functionalities of our Python package, crandas.

Platform functionality

Account setup

• General information
  • Name
  • Email
  • SSO (additional fee)
  • 2FA
• Password update
• Key pair generation
• Language preference (options English or Dutch)

User management

• Add users via email
• Add users via .csv file
• Update role of user
• Remove user
• Available roles
  • Primary Admin
  • Data Provider
  • Analyst
  • Approver
  • Requestee

System

• Download crandas connection file
  • Single connection file
  • Manual (individual files)
• Version visibility
• Legal information
  • Privacy Policy
  • Terms & Conditions

Analysis

• Analysis overview
  • Including name, owner, created at, status and actions (start approval, download and view details)
• Create new analysis
  • General information including name and description
  • Approvers selection from a list of confirmed approvers
  • Analysis in which the user can add a JSON file
• Approval
  • Review the analysis scripts to guarantee that they do not reveal any unnecessary information.

Data request

• Sent overview
  • Overview including name, requester, submit by date, status and link to details
• Create new request
  • General information including name description and soft deadline
  • Requestees selection from list of invited and/or registered users
  • Data specification selection including column name, data type, allowance empty cells, specification and actions
• Received overview
  • Overview including name, requester, submit by date, status and details
• Submit response
  • Information box
  • Data source information form
  • File uploader for .csv including delimiter picker
  • Data validation layout
• Data request view
  • General information including name, requester, description and see specifications button
  • Requestees overview including status, owner, name and handle (when available)
  • Script to copy paste for crandas
  • Option to duplicate and/or delete data source based on role

Data import

• Create new data source
  • General information including name and remarks
  • File type specification including delimiter picker
  • Data validation layout
• Data source detail view
  • General information including name, owner, created at and handle
  • Script to copy paste in crandas
  • Data layout overview
  • Option to duplicate and/or delete data source based on role

Other

• Direct link to help center
• Log-out
• Email notifications
• Graphical profile

Crandas functionality

• Integer
• Fixed point
• Boolean
• Strings
• Dates
• Bytes

• Importing and exporting data:

• • Import data from various sources (through crandas we support many file types as you can read in the data using pandas first).

• Basic table operations:

• • Assign: Create new columns in a dataframe
  • Project: Return table with same rows but a selection of columns
  • Filtering: Select rows based on conditions (e.g., filter rows where age > 30).
  • Slicing: Select rows based on their indices.
  • Shuffling: Return a table with rows randomly shuffled
  • Sample: Take a random subset of the rows.
  • Sorting: Arrange data by specific columns (ascending or descending).
  • Aggregation: Compute summary statistics (e.g., mean, sum, count) for groups of data.

• Working with numeric, text, and date data:

• • Numeric data (integer and fixed point):

• • • Perform mathematical operations (e.g., addition, subtraction) on numeric columns.
    • Calculate descriptive statistics (min, max, mean, median, variance).

• • Text data:

• • • Manipulate text columns (e.g., extract substrings, replace values).
    • Perform case conversions (uppercase, lowercase).
    • Supports substring search (contains) and regular expressions
    • Supports hash functions (currently only RIPEMD160)

• • Date data:

• • • Extract components (year, month, day) from date columns.
    • Calculate time differences.

• Handling missing data:

• • Identify and handle missing values (NaN or NULL).
  • Replace or delete missing data

• Categorical data operations:

• • Quantization: Divide numeric data into intervals or “bins” and map to discrete categories.
  • Groupby: split-apply-combine:

• • • Group data by one or more columns.
    • Apply functions to each group (e.g., compute group-wise averages).
    • Combine results into a new table.

• Merge, join, and concatenate:

• • Merge/join: Combine data from different tables based on common columns (similar to SQL joins).
  • Concatenate: Stack tables vertically or horizontally.

• Machine learning models (both training and evaluation):

• • Binomial logistic regression: Predict binary outcomes (e.g., yes/no, success/failure).
  • Multinomial logistic regression: Predict multiple categories (e.g., classifying animals into species).
  • Ordinal logistic regression: Predict ordered categories (e.g., customer satisfaction levels).
  • Linear regression: Model relationships between variables.
  • K-Nearest neighbors: Predict based on similarity to neighbors.

• Visualisation: Visualise your aggregated output (e.g. using matplotlib or plotly).

---

Thank you for your time to read this article. If you have feedback or if you seek more information on specific topics, leave your comments below or reach out to support@rosemanlabs.com