Welcome to the Jupyter demo

This is an example .ipynb notebook

import sys
print(sys.version)

3.7.9 (default, Oct 19 2020, 15:13:17)
[GCC 7.5.0]

from IPython.display import Image

Image("http://sipi.usc.edu/database/preview/misc/4.2.03.png")

Interactive widgets

Example that render an interactive widget

import ipywidgets as ipw

slider = ipw.IntSlider(min=0, value=10, max=20)
slider

slider

import panel as pn
pn.extension()

def f(x):
    return x

pn.interact(f, x=10)

EGI services

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

data = pd.read_csv('data/datatest.txt')

Dx = data["date"]

data['date'] = data.date.astype('datetime64[ns]')
data = data.set_index('date')

data.tail()

Matplotlib is building the font cache; this may take a moment.

	Temperature	Humidity	Light	CO2	HumidityRatio	Occupancy
date
2015-02-04 10:38:59	24.290000	25.700000	808.0	1150.25	0.004829	1
2015-02-04 10:40:00	24.330000	25.736000	809.8	1129.20	0.004848	1
2015-02-04 10:40:59	24.330000	25.700000	817.0	1125.80	0.004841	1
2015-02-04 10:41:59	24.356667	25.700000	813.0	1123.00	0.004849	1
2015-02-04 10:43:00	24.408333	25.681667	798.0	1124.00	0.004860	1

from matplotlib.figure import Figure
from matplotlib.backends.backend_agg import FigureCanvas
%matplotlib inline

plt.figure(figsize=(10,5));
plt.plot(data.CO2);
plt.xticks(rotation=45);
plt.xlabel('Date', fontsize = 20)
plt.ylabel('Temperature',fontsize = 20)

plt.show()

Můžu použít nějakou statistiku - třeba rolling mean:

\[\tau = \frac{\varpi - \omega}{\sqrt{\frac{\sum \varpi - \omega}{N(N-1)}}}\]

Tohle sice není rolling mean, ale aspoň vidíte, že se sem dá dát i LaTeX

def mpl_plot(avg, highlight):
    fig = Figure()
    FigureCanvas(fig)
    ax = fig.add_subplot()
    avg.plot(ax=ax)
    if len(highlight): highlight.plot(style='o', ax=ax)
    return fig

def find_outliers(variable='Temperature', window=50, sigma=10, view_fn=mpl_plot):
    avg = data[variable].rolling(window=window).mean()
    residual = data[variable] - avg
    std = residual.rolling(window=window).std()
    outliers = (np.abs(residual) > std * sigma)
    return view_fn(avg, avg[outliers])

find_outliers(variable='CO2', window=50, sigma=10)

import panel as pn
pn.extension()

pn.interact(find_outliers)