CHEMISTRY / 01–06

Chemistry

6 independent, runnable lessons connecting questions, traceable inputs, transparent methods and interpretation. One notebook download starts each workflow.

Learning path 01 → 02 → 03 → 04 → 05 → 06; every lesson also runs independently.
Download this discipline in both languages

6 recipes

Reconcile molecular formulas and calculated properties — executed figure
01 / Introductory50 minOffline core

Reconcile molecular formulas and calculated properties

Parse eight PubChem formulas, reconstruct approximate molar masses and compare elemental mass fractions and reported descriptors.

numpy · pandas · matplotlib · IPython
Read standardized molecules as graphs — executed figure
02 / Intermediate50 minOffline core

Read standardized molecules as graphs

Parse real V2000 records, preserve formal charges, compare connected components and cycles, and export inspectable graphs.

numpy · pandas · matplotlib · networkx · IPython
Inspect a calculated caffeine conformer — executed figure
03 / Intermediate50 minOffline core

Inspect a calculated caffeine conformer

Preview the PubChem 3D conformer offline, compare bond lengths and audit heavy-atom distance geometry.

numpy · pandas · scipy · matplotlib · py3Dmol · IPython
Solve an ideal weak-acid titration — executed figure
04 / Intermediate50 minOffline core

Solve an ideal weak-acid titration

Use mass and charge balance across a strong-base titration, compare the buffer approximation and inspect species fractions.

numpy · pandas · scipy · matplotlib · IPython
Evaluate a reference vapor-pressure correlation — executed figure
05 / Intermediate50 minOffline core

Evaluate a reference vapor-pressure correlation

Reproduce IAPWS reference pressure in its valid range, check reference points and examine a Clausius–Clapeyron approximation.

numpy · pandas · matplotlib · IPython
Audit a simulated absorbance calibration — executed figure
06 / Intermediate60 minOffline core

Audit a simulated absorbance calibration

Generate declared Beer–Lambert teaching observations, fit a calibration, inspect residuals and propagate uncertainty to an unknown.

numpy · pandas · matplotlib · IPython