2023-Q3-AI 1.Intro, Python Basics (OOP)

1.1. Video / Materials

Video (18 Jul 2023, 10:00): https://youtube.com/live/AdBJ8McH57o?feature=share

Jamboard: https://jamboard.google.com/d/1x4xjRdHMGPoIVISfe5LGY7Wuj2cJ5sFQDeOC2cviRIc/viewer?f=0

Jamboard ir iedotas edit tiesības Lekciju nepieciešams ierakstīt ar OBS, screen recording, Youtube live streaming key: ea11-mrgb-4jg2-4ajc-d4hr

Par katru uzdevumu dodam 100 punktus

Paraugs - iepriekšējā gada video un jamboard

Video: https://youtube.com/live/HtMV4LAujZE?feature=share

Jamboard: https://jamboard.google.com/d/1tKWa9a_Zu7UrGMAShRCUNNht6XetLnCB6fDCTgtKESU/edit?usp=sharing

1.2 Prepare environment

Before the course it is necessary to complete following steps:

Install and setup PyCharm IDE (you can access student version with full features as well): https://www.jetbrains.com/pycharm/
Install Anaconda 3.9 Python environment: https://www.anaconda.com/products/individual
Install PyTorch library, using conda package management system (For every Operating System you will have command generated at: https://pytorch.org). Command to install PyTorch without CUDA/GPU support would be: conda install pytorch torchvision torchaudio -c pytorch
Install sklearn: conda install scikit-learn Install matplotlib: conda install matplotlib
Run tempalate hello_world.py (below task), using PyCharm IDE (If all is successful then you will not have errors and all libraries need for work would be functional)
Submit screenshot of running output of PyCharm IDE to complete this task.

More information about conda environments:

Each conda environment contains a set of libraries that are available only in that environment
Create new environment: conda create --name your_name
Activate environemnt: conda activate your_name
Only then install libraries when conda environment is activated.
Set project interpreter as your conda environment in PyCharm IDE settings
More about Conda Environments: https://towardsdatascience.com/getting-started-with-python-environments-using-conda-32e9f2779307 To ease work if you connect to sessions rem

Also recommended materials:

Complete tutorial: https://www.w3schools.com/python/
Implement Python and carefully read this tutorial: http://karpathy.github.io/neuralnets/
Watch a video course from Stanford on deep learning: https://www.youtube.com/watch?v=NfnWJUyUJYU&list=PLkt2uSq6rBVctENoVBg1TpCC7OQi31AlC
Start reading an online book about deep learning: http://neuralnetworksanddeeplearning.com
Review advanced mathematics, derivatives, differential equations, and linear algebra.

Template: http://share.yellowrobot.xyz/quick/2023-1-12-DCFFD523-CB00-4082-8CBA-7867E7325C85.zip Result: http://share.yellowrobot.xyz/upic/3908a85279643b5bb76c1e6f2de8291e_1689350893.jpg

1.3 Implement function and submit source code

$f(x, b)$ . Do not use pre-built functions (online). Do not use numpy or other libraries, only pure Python code.

Submit the source code and a screenshot. Bonus points if you use recursion.

Create code for the following equation.

http://share.yellowrobot.xyz/upic/df7e6c9b6b8f880de6ee2d3e5b740e75_1673883820.png

\begin{matrix} (2) & f (x, b) = (x^{b} + b) \cdot (x^{b - 1} + b - 1) \cdot (x^{b - 2} + b - 2) \cdot \dots \cdot (x^{0} + 0) \end{matrix}

1.4. Implementēt function and display visually using matplotlib

Submit source code and screenshot, additional bonus if recursion is used. Extra points if numpy is used.

Equation: http://share.yellowrobot.xyz/quick/2023-1-16-E7F9E2D0-82B4-4C10-8DB5-5EC0CEF84784.png

Result: http://share.yellowrobot.xyz/upic/596fdf6bed2c4bc1c1ef6476ec72500c_1673884738.png

\begin{matrix} (3) & f (x) = \sin (2 x) + 2 e^{3 x} \end{matrix}

plot f(x) = \sin(2x) + 2e^{3x}

1.5. Implement Python object-oriented code using UML class diagrams.

Implement Python object-oriented code using UML class diagrams.

vec_pos and others are initialized as 2D numpy vectors np.array([0., 0.]).

All functions initially do nothing; the scope code only contains pass.

Only the class structure needs to be created; submit a Python file.

Optional: Bonus points if you implement the Singleton pattern inside the static functions Game.get_instance() and Player.get_instance() (Information about design patterns can be found at https://refactoring.guru/design-patterns)

Template:

http://share.yellowrobot.xyz/quick/2023-1-16-E06CA8D8-A955-4248-BE8A-2410B04A7654.zip

Schema:

http://share.yellowrobot.xyz/quick/2023-1-16-BBCEFB14-D351-423E-8C20-19DD5A888995.jpg

Materials

print function
basic data types - meaningful variable names - viss, kas notiek veikalā (produktu cenas, svari utt)
Apskatīties pēc kārtas katru sadaļu - https://www.w3schools.com/python/default.asp
mathematical operations +,-…, += syntax, round, abs
casting basic data types
if, elif, else, and, or
for, while loops
functions def
classes, inheritance, encapsulation, polymorphism

😊 Pointer example


xxxxxxxxxx
8
1
shelf_1 = Shelf()
2
shelf_2 = Shelf()
3
shelf_2.items = ['bread', 'milk']
4
shelf_1 = shelf_2
5
shelf_1.items.append('butter')
6

7
print(shelf_1.items)
8
print(shelf_2.items)

complex data types - list, dict, tuple
dict functions “exist in”, del, set value, get value
list functions len, append, index, __del__ un del keyword
loop for
list access elements by index
a[-0:] => -0 == 0 😊 a[-1] # last item in the array (jābūt IF condition > 0) a[-2:] # last two items in the array (jābūt IF condition > 0) a[:-2] # everything except the last two items

😊 Pointer example


xxxxxxxxxx
5
1
night_shift = [1, 2, 3]
2
morning_shift = night_shift
3
morning_shift.append(4)
4
print(night_shift)
5
print(morning_shift)

tuple - () problem
library import numpy
numpy complex data type
😊 numpy vs list [1] * 4
😊 numpy + vs list + [1, 1] + [2, 2]
numpy concat vs list concat
np.arange(len(y))
np.dot
np.transpose
1.3 funkcijas implementācija
matplotlib
plot 1.3
- plot + show
  - plot strypes => fmt parameters (https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.plot.html)
  - r- => red line
  - ro => red points
  - rs => red square
  - r-- => red doted
  - r^ => red triangles
- plt.plot(x, y, linewidth=2.0)
- plt.ylabel('axis')
- plt.yscale('log') plt.yscale('linear')
- plt.axis(['a', 'b', 'c', 'd'])
- plt.title('Histogram of IQ')
- $\mu=100,\ \sigma=15$ ')
- plt.grid(True)
- plt.subplot(221) (syntax nrows, ncols, index)
UML / OOP