git branch. This will tell you in which branch you are currently working locally. Because there is no other branches, you should see master.git branch [your-name-feature-name], where your-name is your first name or user name on Github.com, and feature-name is the feature you are working on in the branch. For this exercise feature-name can be called first-PR. git checkout [your-name-feature-name].git add sandbox.ipynb then commit changes to the branch git commit -a -m sandbox.ipynb or by uploading the file directly to your branch on Github,com, and push the branch to Github: git push -u sandbox.ipynb. The option -u is required the first time you push.Once you are ready to push the changes in your branch to the master repository, you need to create a pull request for the team to review your code before integration.
Never merge your own pull request to the master. Somebody else in your team (ideally everyone as a group) should review your pull request and agree before it is merged.
To bring the changes to the main repository, one of the team members (who is not the one making the pull request) can merge the branch into the master branch.
Since your branch is now closed, you won't be able to push any further changes to the remote repository.
git checkout mastergit pull.You should communicate extensively as a team, to avoid working on the same portions of the same files. Editing the same cells concurrently will result in conflicts.
Because everyone will be working in parallel, possibly on the same notebook, you need to make sure that your local version (master and branch) integrates the latest updates to the remote master.
git fetch command allows your local repository (master and branches) to be aware of the changes to all the branches in the repository (including master).git merge. For example, to include the update changes from the master branch to your branch: git merge origin/master. This will allow you to address any conflicts that your local changes may cause to the common, remote repository. ## Person 1 should add a cell before this code cell ##
from ipywidgets import interact
import ipywidgets as widgets
def f(x):
return x
interact(f, x=10) # integer argument generates an IntSlider widget
## Person 2 should add a cell before this code cell ##
interact(f, x=True) # boolean argument generates a Checkbox widget
## Person 3 should add a cell before this code cell ##
interact(f, x='Text') # text argument generates a Text widget
## If your team has 4 members, person 4 should add a cell before this code cell ##
def f(a, b):
return print('a + b =', a + b)
interact(f, a=1, b=1);
In this part, we will get an introduction to IPython widgets. These are tools that allow us to build interactivity into our notebooks often with a single line of code. These widgets are very useful for data exploration and analysis, for example, selecting certain data or updating charts. In effect, Widgets allow you to make Jupyter Notebooks into an interactive dashboard instead of a static document.
Widgets are eventful python objects that have a representation in the browser, often as a control like a slider, textbox, etc. For example, allow the user to move a slider that results in another value being updated.
Ref: ipywidgets docs
Let's load the list of top 10 movies that we had scraped from the criterion website.
import pandas as pd
df = pd.read_csv('top10-lists.csv')
df.head()
Now, let's create a slider widget that will allow us to interactively select the number of rows to be displayed.
def frows(x):
return df.head(x)
interact(frows, x = 10);
Playing with the slider, you can note that the range of possible values is not appropriate. Let's constrain the slider to the range [0; number of rows]. For that, we can pass the IntSlider as the keyword argument for x, and set the parameters as desired.
interact(frows, x=widgets.IntSlider(min=0, max=df.shape[0], value=3));
Now, let's filter the data, and plot only the selected column, by column ID.
def datacol(colnum):
return df.loc[:,list(df)[colnum]].head(n=3)
interact(datacol, colnum=widgets.IntSlider(min=0, max=len(df.columns)-1, value=0));
The column ID is not very informative. A better selection widget would be to let the viewer select the desired column from a drop-down list showing the column names.
def datacol(colname):
return df[colname].head(3)
interact(datacol, colname = list(df)); # a dropdown is generated
# if a list of tuples is given
Here is an example where the viewer can select a movie name from the list of movies in our dataframe, without the duplicates.
import pandas as pd
import numpy as np
# dropdown menu of titles, removing the duplicates, and sorting
dd = widgets.Dropdown(options = df.title.drop_duplicates().sort_values())
# display dropdown and dataframe
display(dd)
interactive, interact_manual, continuous_update, and interactive_output all provide additional flexibility to interact. In the example below, we create different sliders to manipulate different t distributions.
from ipywidgets import interact
import ipywidgets as widgets
from scipy.stats import t
import matplotlib.pyplot as plt
import numpy as np
def plott(mean,sd, df):
fig, ax = plt.subplots()
x0 = np.linspace(t.ppf(0.01, df = 1, loc = 0, scale = 1),t.ppf(0.99,df=1, loc = 0, scale = 1), 100)
x = np.linspace(t.ppf(0.01, df = df, loc = mean, scale = sd),t.ppf(0.99, df=df, loc = mean, scale = sd), 100)
plt.plot(x0, t.pdf(x0, df=1, loc = 0, scale =1), lw=5, alpha=0.6, label=r'$t_{1}$')
plt.legend()
plt.plot(x, t.pdf(x, df=df, loc = mean, scale =sd), lw=5, alpha=0.6, label=r'$t_{\nu}$')
plt.legend()
plt.axvline(x=0, color = 'red')
plt.axvline(x=mean)
#plt.show()
interact(plott, mean = widgets.IntSlider(value=1, min=0, max=20, step=1),
sd = widgets.IntSlider(value=1, min=1, max=5, step=1),
df = widgets.IntSlider(value=1, min=1, max=30, step=1))
widgets.HTML(
value="<b>Drag the slider to explore different t distributions</b>",
)
Exploration of normal distributions
from ipywidgets import interact
import ipywidgets as widgets
from scipy.stats import norm
import matplotlib.pyplot as plt
import numpy as np
def plotnorm(mean,sd):
fig, ax = plt.subplots()
x0 = np.linspace(norm.ppf(0.01, loc = 0, scale = 1),norm.ppf(0.99, loc = 0, scale = 1), 100)
x = np.linspace(norm.ppf(0.01, loc = mean, scale = sd),norm.ppf(0.99, loc = mean, scale = sd), 100)
plt.plot(x0, norm.pdf(x0, loc = 0, scale =1), lw=5, alpha=0.6, label='N(0,1)')
plt.legend()
plt.plot(x, norm.pdf(x, loc = mean, scale =sd), lw=5, alpha=0.6, label=r'N($\mu, \sigma^2)$')
plt.legend()
plt.axvline(x=0, color = 'red')
plt.axvline(x=mean)
#plt.show()
interact(plotnorm, mean = widgets.IntSlider(value=1, min=0, max=7, step=1),
sd = widgets.IntSlider(value=1, min=1, max=5, step=1))
widgets.HTML(
value="<b>Drag the slider to explore different normal distributions</b>",
)
interactive Exercise¶Create an interactive scatter plot that shows the effect of mean, variance, and sample size on a fitted simple linear regression line. Assume that the independent variable has $N(\mu,\sigma^2)$ distribution.
## Put your code here
The Widgets-Overview.ipynb notebook contains additional examples of interactive widgets that you can bear inspiration from.
You can visualize your notebook as a webpage using voila.
From your notebook, click on the "Voila" icon to open a pop-up window of your notebook rendered as a webpage.
You can also type voila sandbox.ipynb from your terminal.
The NuminaSimpleQueries.ipynb notebook includes examples of simple queries to the Numina API.
Exercise: Extend the code in the "Counts" section to plot the timeseries of the count of pedestrians.