udacity/aipnd

Code and associated files for the AI Programming with Python Nanodegree Program

AIPND

This repository contains code and associated files for the AI Programming with Python Nanodegree program. This repository consists of a number of tutorial notebooks for various coding exercises and programming labs that will be used to supplement the lessons of the course.

Table Of Contents

Tutorial Notebooks

• Linear Algebra Essentials
  • Vectors Lab Learn how to graph 2D vectors.
  • Linear Combination Lab Learn how to computationally determine a vector's span and solve a simple system of equations.
  • Linear Mapping Learn how to solve some problems computationally using vectors and matrices.

Programming Labs

• Intro to Python Lab - Classifying Images Lab: Determine which CNN architecture model works best at classifying images of dogs and their breeds.

NumPy and Pandas Mini-Projects

• NumPy Mini-Project Mean normalize and separate data
• Pandas Mini-Project Get statistics from stock data

Matplotlib

• Matplotlib Notebooks containing practice exercises for the Matplotlib lesson(s)

Quiz Notes

• Notes: This directory contains more information about certain quizzes that are testing more challenging concepts. The filename indicates which quiz the note is covering. Additionally, one will find the Frequently Asked Questions for the Intro to Python Lab. Click on the filename to view the contents of the notes on a quiz or the Intro to Python Lab.

Dependencies

Each directory has a requirements.txt describing the minimal dependencies required to run the notebooks in that directory.

pip

To install these dependencies with pip, you can issue pip3 install -r requirements.txt.

aipnd's Issues

pandas mini project solution mistake

in the panda mini project's solution , in the input7 error

calculating X_test and X_crossVal is wrong as it takes 80%.

Missing Matplotlib Dataset

A file is missing from the Matplotlib / data directory. A dataset is consistently used in the lessons that includes "Alpha","Beta", "Gamma", etc. values.

In order to properly do those exercises, I had to randomly generate Grade data as a substitute:

# Back to auto-generated data for the following steps
student_grades = []
for student in range(30):
  student_grades.append(random.choice(['A','B','C','D','F']))
grades = pd.DataFrame({'Grades': pd.Series(student_grades)})
grades.head()

I had to replace all instances of df in his example code with grades, and replaced all instances of cat_var with Grades. His examples in messing with the resulting bar graphs should apply in the same way with grades as he is doing with Alpha, Beta, etc.

This worked properly with examples, such as categorical ordering:

# With grades, or categorical / ordinal data that is ranked by type,
# we may care about an explicit order. Ordering it up!
level_order = ['A', 'B', 'C', 'D', 'F']
ordered_cat = pd.api.types.CategoricalDtype(ordered = True, categories = level_order)
grades['Grades'] = grades['Grades'].astype(ordered_cat)
# Display graph
base_color = sb.color_palette()[0]
sb.countplot(data = grades, x = 'Grades', color = base_color);

I have added more detailed information as an answer to a question in the Udacity Knowledge Base, where someone ran into the same problem, accessible behind an authenticated page: How to practice examples (matplotlib/seaborn) with no .csv file available for download??

Extra usage of variable 'found'

The usage of variable 'found' starting from line 148 is not required and can be done without it. Usage of 'found' is creating unboundLocalError.

Error Message when trying to open in Jupyter Notebook

Hello, I downloaded Mean Normalization and Data Separation.ipynb from https://github.com/udacity/AIPND/tree/master/NumPy%20Mini-Project, but the following error message appeared when I tried to open the file in a Jupyter Notebook:

"Unreadable Notebook: C:\Users\Christophe\Documents\Informatique\MOOC\Data Scientist Nanodegree - Udacity\Jupyter Notebooks\Mean Normalization and Data Separation.ipynb NotJSONError("Notebook does not appear to be JSON: '\n\n\n\n\n\n\n<html lang...")"

Any suggestions? Thanks.