For scientific research

Example workflows

Task	How to do it
Experimental analysis	”Run a paired t-test comparing pre and post treatment measurements”
Data visualization	”Create a scatter plot with regression line and 95% confidence bands”
Literature data	Use deep research to “Find recent studies on [topic] and summarize key findings”
Power analysis	”Calculate the required sample size for detecting a medium effect with 80% power”
Meta-analysis	”Calculate pooled effect sizes from these study results using random effects model”
Data cleaning	”Standardize units, handle missing values, and flag impossible measurements”

Statistical analysis — t-tests, ANOVA, chi-square, regression, non-parametric tests
Feature engineering — Data transformation and normalization
Deep research — Literature search and synthesis
Python — Full access to SciPy, StatsModels, scikit-learn
Import/export — Handle CSV, XLSX, JSON, Parquet, and more