Positron: first contact

Positron is a data-science IDE built for Python and R. It brings a professional code editor together with interactive consoles, variables, plots, notebooks, Quarto and a Data Explorer, so exploratory work and a saved analytical record can stay together.

Coding Polyglot: Python is the primary language in this on-ramp, but Positron is deliberately not a Python-only tool. Statisticians can use first-class R support in the same workspace, while Code OSS extensions add languages such as SQL, JavaScript, C++ or Rust when a project needs them.

The safe mental model is data + saved code → session objects + inspectable output. The Console and Variables pane are live working memory; the saved script is the reproducible record.

Positron Desktop is available for Windows, macOS and Linux. Use an institutionally managed installation if one is provided; otherwise download the current desktop release from Posit.

1. Install and open Positron. Use the official Positron download and allow its normal update checks.
2. Open one project folder. Keep the teaching CSV and Python script together so a relative filename remains reproducible.
3. Select a Python interpreter. Use the interpreter selector in the Top Bar. Positron can help install Python if no supported interpreter is available.
4. Start a Python session. Confirm that the selected interpreter appears beside the Console before running the script.

Prefer R? Select an R interpreter instead and use the supplied R script without leaving Positron. The editor, Variables pane, plots and Data Explorer serve both languages.

Put the supplied .py file and CSV in one folder, then open that folder in Positron. Open the Python file and run it with the Run action or Ctrl+Enter on Windows/Linux and Cmd+Enter on macOS.

This block uses pandas to create a data frame named cohort, check its shape and answer the same medication- review and age-summary questions as the other on-ramps. Select cohort in Variables afterwards to open it in the Data Explorer.

import pandas as pd

cohort = pd.read_csv("gerostats_medication_review_cohort.csv")
counts = cohort["med_review"].value_counts().sort_index()

print("shape:", cohort.shape)
print("medication review counts:")
print(counts.to_string())
print("medication review percentages:")
print((counts / len(cohort) * 100).round(1).to_string())
print("age summary:")
print(cohort["age_years"].agg(["mean", "min", "max"]).round(2).to_string())

shape: (720, 16)
medication review counts:
med_review
0    517
1    203
medication review percentages:
med_review
0    71.8
1    28.2
age summary:
mean    78.35
min     65.00
max     96.00

Values 0 and 1 are codes, not yet reader-friendly labels. The display confirms that Python read the same 720-row fictional cohort used by the other software desks; it does not test whether medication review changed an outcome.

Terms you now need

Workspace

the folder and files opened together as one project context.

Interpreter

the selected Python or R installation that executes code.

Session

the live language process containing current objects, packages and console history.

Data Explorer

a temporary interactive view of raw files or in-memory data frames, with filters and summaries.