First run

Before setup: use readiness and Python Foundation, workspace acquisition and environment choices. Python Foundation is preparation, not a formal admission prerequisite. Our Python-to-ML bridge adds file and table skills.

Obtain a checkout or extract the complete source archive, then open a terminal in its ml_ai_it folder. Use Python 3.11, 3.12 or 3.13; CI checks the core modules on 3.12 (Windows and Linux) and 3.13 (Windows). If Python is missing, use the official download page and your platform's installer instructions. Check your version before proceeding.

Commands below are selectable text. py is the usual Windows Python launcher; if it is unavailable but python --version reports the intended interpreter, use python instead. On macOS/Linux, use python3. Report an unavailable command rather than entering unrelated installation commands suggested by an unverified source.

Module 1: no package installation needed

Windows PowerShell:

py --version
py examples/first-run/scripts/first_run.py --storage-percent 85 --output runs/module01/first-run.txt

macOS or Linux:

python3 --version
python3 examples/first-run/scripts/first_run.py --storage-percent 85 --output runs/module01/first-run.txt

Expect storage used 85 percent, threshold 80 percent, and Review storage. Open the saved text file. This is a human-written rule; no model is trained. Continue with Module 1.

Module 2: read CSV and JSON

Windows PowerShell:

py examples/service-files/scripts/summarize.py --tickets data/samples/service-files/tickets.csv --events data/samples/service-files/events.json --output runs/module02

macOS or Linux:

python3 examples/service-files/scripts/summarize.py --tickets data/samples/service-files/tickets.csv --events data/samples/service-files/events.json --output runs/module02

Expect 4 tickets, 60 total minutes, average 15 minutes, and 3 events. Inspect summary.txt, report.json, and category_summary.csv under runs/module02/. See Module 2 for interpretation and input-error practice.

Module 3: add the data libraries

Create an isolated environment and install the local package. Calling the environment's Python directly avoids a script-activation requirement. These commands install third-party dependencies; internet access is normally needed. The constraints record the NumPy/pandas versions used in development.

Windows PowerShell:

py -m venv .venv
.\.venv\Scripts\python.exe -m pip install -e . -c environments/constraints-tested.txt
.\.venv\Scripts\python.exe examples/network-events/scripts/prepare.py --input data/samples/network-events/events.csv --output runs/module03

macOS or Linux:

python3 -m venv .venv
.venv/bin/python -m pip install -e . -c environments/constraints-tested.txt
.venv/bin/python examples/network-events/scripts/prepare.py --input data/samples/network-events/events.csv --output runs/module03

py -m venv uses your default Python; run py --list if you have several versions, and record the version when reporting an issue. Expect 12 input records, 6 accepted, 6 rejected, and 1 accepted record retaining missing latency. Observed latency mean/median are 24/20 milliseconds. See Module 3.

Module 4: add plotting libraries

After creating .venv, install the visualization extra and run exploration. In Windows PowerShell:

.\.venv\Scripts\python.exe -m pip install -e ".[visualization]" -c environments/constraints-tested.txt
.\.venv\Scripts\python.exe examples/service-workload/scripts/explore.py --input data/samples/service-workload/daily.csv --output runs/module04

On macOS/Linux:

.venv/bin/python -m pip install -e ".[visualization]" -c environments/constraints-tested.txt
.venv/bin/python examples/service-workload/scripts/explore.py --input data/samples/service-workload/daily.csv --output runs/module04

Expect 24 distinct observations, 4,180 requests, 23 observed latencies and one unavailable latency. Open runs/module04/index.html for the five figures with actual tables/descriptions, or use its plain-text/CSV evidence. See Module 4 for preprocessing and interpretation. No display server is required for generating figures.

Module 5: regression and real public data

After creating the environment, add the regression/visualization extras, download the attributed UCI data, and run a fresh study. Windows PowerShell:

.\.venv\Scripts\python.exe -m pip install -e ".[regression,visualization]" -c environments/constraints-tested.txt
.\.venv\Scripts\python.exe examples/bike-demand/scripts/study.py download --run runs/datasets/bike-sharing
.\.venv\Scripts\python.exe examples/bike-demand/scripts/study.py all --input runs/datasets/bike-sharing/day.csv --run runs/module05 --holdout-days 110 --plots --dataset-label "UCI Bike Sharing daily counts, CC BY 4.0"

macOS/Linux:

.venv/bin/python -m pip install -e ".[regression,visualization]" -c environments/constraints-tested.txt
.venv/bin/python examples/bike-demand/scripts/study.py download --run runs/datasets/bike-sharing
.venv/bin/python examples/bike-demand/scripts/study.py all --input runs/datasets/bike-sharing/day.csv --run runs/module05 --holdout-days 110 --plots --dataset-label "UCI Bike Sharing daily counts, CC BY 4.0"

Expect 504 training, 110 validation, and 110 final-test examples after seven warm-up days. Open runs/module05/index.html or its text/CSV evidence. Use a new study directory for a replay. Module 5 explains separate stages, the forecast origin, source/license, and the clearly synthetic offline alternative. The first real-data download needs internet, while later runs reuse a checksum-verified cache.

Notebook option

Notebooks are optional execution companions. Every course notebook, including the Module 1 and 2 practice notebooks, needs the core environment: the practice checks use NumPy even though the Module 1 and 2 scripts do not. Do these steps once, from the ml_ai_it folder.

Create .venv and install the package with the Module 3 commands above.
Install a notebook interface into that environment and register it as the course kernel.
Launch the interface from the same folder.

Windows PowerShell:

.\.venv\Scripts\python.exe -m pip install notebook ipykernel
.\.venv\Scripts\python.exe -m ipykernel install --user --name ml-ai-core --display-name "Python (course core)"
.\.venv\Scripts\python.exe -m notebook

macOS or Linux:

.venv/bin/python -m pip install notebook ipykernel
.venv/bin/python -m ipykernel install --user --name ml-ai-core --display-name "Python (course core)"
.venv/bin/python -m notebook

Course notebooks request the kernel named Python (course core). If the interface asks you to choose a kernel, or reports that ml-ai-core is missing, choose Python (course core) (Kernel → Change kernel), or rerun the registration command. Framework notebooks (Modules 8–12 and 14) request Python (course frameworks), registered by the framework setup.

From the opened interface, open the module's notebook under notebooks/lessons/, or its practice notebook under notebooks/practice/. Select and run code cells in order. Save with the interface's File → Save command, restart the kernel, and run all cells. The notebook imports shared source from the complete repository folder; downloading a notebook alone is insufficient. Use the documented script/text route when it better fits your interface needs; the accessibility guide records interface testing scope. See setup and accessibility status.

Save and troubleshoot

Use different output names/directories for experiments; rerunning replaces generated files at the same paths. Keep raw samples unchanged. A failed run may leave an older report from a previous successful run, so retain commands and error text with your experiment.

Command unavailable: check the installed Python command and version.
File not found: check the working folder and the exact input path.
Missing columns or invalid value: inspect the CSV header and named field; use a synthetic copy for repair practice.
Missing NumPy/pandas: install with the same environment executable used to run Module 3.

For help, retain the command, Python/platform version, error text, and small synthetic input. Screenshots are not required. Keep credentials and personal information out of public issues.

Modules 6–14

Choose a module from the curriculum map. Modules 6, 7, and 13 use the core environment; Modules 8–12 and 14 use the isolated CPU framework setup. Each entry supplies its script, commented notebook, actual source, and evidence requirements. Use fresh output directories and retain verified data/model caches for replay.

Maintainer verification

To execute the curated notebooks programmatically, install the verification extra in the relevant environment, register its Jupyter kernel, and follow development verification. A notebook frontend alone is not the complete automated-check setup. The framework guide uses its own environment and kernel. Clean source notebooks and saved executed build copies have different purposes.

View source for this page on GitHub