Formation - Python for Data Science

MUTUALISÉE

Intra-entreprise

Référence : 337QT2xmzQaW

Durée : 21 h sur 3 j

À partir de

525 €

Pré-inscription

Elara Formations

ANTIBES

Mutualisée

Une formation mutualisée est une session organisée initialement pour les salariés d’une même entreprise. Mutualiser une formation consiste à l’ouvrir à des participants externes dans le but d’optimiser le taux de remplissage et de partager la facture au prorata. Le tarif de la session est forfaitaire quel que soit le nombre de participants.

Intra-Entreprise

Une formation intra-entreprise réunit dans une même session les salariés d'une seule entreprise. Elle se déroule généralement dans les locaux de l'entreprise. Le tarif est convenu pour la formation d'un groupe de salarié.

This training aims to provide you with a solid foundation in Python data science programming. We will cover the modern usage, subtleties, and pitfalls of NumPy, pandas, and Seaborn for visualization.

Détails de la formation

Méthodes et outils pédagogiques

Presentation of the concept using examples followed by exercises.

Objectifs de la formation

This training covers the fundamentals necessary for any data analysis work in Python. In particular, we will explore the NumPy, pandas, and Seaborn libraries. Throughout the training, we will use Jupyter notebooks. Special attention will be paid to the most recent and effective techniques for data manipulation.

Méthodes d'évaluation

There is no evaluation.

Les plus

This training is made by Arnaud Legout a research scientist specialized in complex data analysis using Python. He has been using Python for 20 years in his research activities and has trained hundreds of students and professionals in this language. He is also the co-author of the most popular French-language MOOC on the Python language with more than 150,000 registrant. This MOOC has been widely praised for its educational quality.

Pré-requis

This training is aimed at engineers, doctoral students, post-doc, and researchers who already followed the training session "Introduction to Python in English (level 1)" or with intermediate level in Python (knowing how to write functions and pass arguments to them, understanding iteration in Python, and understanding namespaces and module importation).

Public cible

Tous publics

Sessions programmées

Du 20 mai 2025

Au 22 mai 2025

Pour 8 inscrits

525 €

Pour 4 inscrit

1 050 € HT

(Mutualisée)

2/8 inscrits

Inscrits

Prix part.

1 050 €

840 €

700 €

600 €

525 €

Votre formateur

Arnaud Legout

Arnaud Legout is a research director at Inria with over 20 years of teaching experience at university and corporate training. He has been using Python for 20 years in his research activities and has trained hundreds of students and professionals in this language. Together with Thierry Parmentelat, he created the first French-language MOOC on the Python language in 2014. This MOOC has had over 150,000 registrants since its first edition in 2014 and has been widely praised for its educational quality. This MOOC is used by the mathematics degree program at UPMC and by CentraleSupelec Paris to train their students in programming.

Détails de la session

Lieu de la session :

En distanciel

Modalités d'enseignement :

En distanciel

Langue :

Anglais

Horaires :

9h à 12h30 et 14h00 à 17h30

Possibilité de restauration à proximité :

Non

Restauration incluse :

Non

Informations complémentaires :

En distanciel et en anglais

Programme

Python Review

We will cover how to install your data science Python environment using pip and conda, how to use IPython and Jupyter notebooks, which Integrated Development Environment (IDE) to use (such as VSCode, PyCharm, Spyder, etc.), how to use LLM-assisted programming tools like Copilot, and why to use Git.

NumPy

We will cover the main concepts you need to master to make the most of NumPy:

- The ndarray type
- Reshaping, slicing, and indexing
- Vectorization
- Broadcasting
- Memory representation of ndarray and code optimizations

We will also discuss the subtleties and pitfalls of NumPy and demonstrate how to write NumPy code that can be up to two orders of magnitude faster than regular Python code.

pandas

We will cover all aspects of pandas, including its modern usage, subtleties, and pitfalls. Specifically, we will address:

- Series and DataFrame data types
- Indexes and indexing using .loc and .iloc
- Import and export formats
- Memory representation of Series and DataFrame, Copy on Write (CoW), and code optimization
- Extended dtypes, including pyarrow types
- Method chaining
- Handling missing data
- String operations
- Categorical dtype
- MultiIndexes
- Advanced operations: concat, merge, groupby, and reshaping
- Time series

Seaborn

We will cover the important concepts of tidy data representation, the grammar of graphics, and how to perform data exploration with up to seven dimensions using simple 2D representations for:

- Relational data
- Categorical data
- Distributions

Pré-inscription