git clone https://github.com/ijstokes/wakari_tutorial.gitwakari_tutorial directoryIntroduction to Data Analysis with IPython and Wakari.ipynbSHIFT-ENTER in each cell.Each cell can contain one or more Python statements
In [ ]:
print "This is a cell"
In [ ]:
x = 4
y = 7
result = x + y
print result
In [ ]:
import time
time.asctime()
In [ ]:
print "Seconds since the epoch:", time.time()
print "Years since the epoch (approx): %.2f" % (time.time()/(365 * 24 * 60 * 60))
print "Previously computed value result: ", result
time cells and watch the time change
`A running notebook (like this one, if you've opened it in IPython Notebook") has a Python kernel associated with it, and executing a cell is like running the statements in that cell inside an interactive IPython interpreter.
You can re-order cells and then re-execute them, but re-ordering does not re-run all other cells. For example, try moving the above "seconds since the epoch" cell up near the First Steps section using the ⬆ button in the toolbar, then re-run it.
Notice that the time module and the result variable are still in scope.
When you save your Notebook (with the 💾 button in the toobar) all your input and output cells are saved as well. Try it now, then select File→Close and Halt -- you will get a warning message that the kernel has stopped.
Close the notebook by clicking the ⊗ in the tab.
Re-open the notebook by double clicking on it in the left sidebar menu.
Notice the output has been saved -- this hasn't re-executed. Clear the output by selecting Cell→All Output→Clear.
Now try re-executing all the cells in the notebook by selecting Cell→Run All.
If you moved the time cell up before saving, then you'll get an exception when this runs, since import time hasn't been called yet.
In [ ]:
wont_work = 73 + 'some string'
Notice that Run All will stop if it hits a code cell that throws an exception. Fix 'some string' by making it a number, then select Cell→Run All Below.
You can also delete and insert cells. Select the cell above and then cut it out with the ✂ menu bar button. Add a new cell above this one using the ⤒ button. Try adding some code and executing it.
You can also copy cells with the copy button (right of the scissors), and paste them with the paste button (clipboard two-right of the scissors). Try copying the "time since epoch" cell.
You can use the tool bar drop-down menu to change a particular cell from code to something that will allow you to enter plain text, HTML, or Markdown.
This can even include mathematical notation via LaTeX $$c = \sqrt{a^2 + b^2}$$
Click on any text cell in this notebook to see examples of this.
Markdown is a great way to quickly write readable text that can also render to HTML for the web. Here is a short tutorial
The loop below will take a long time to run. In IPython Notebook we can interrupt the kernel similar to a CTRL-C signal by clicking on the stop button "■" in the menu bar. Selecting Kernel→Interrupt will do the same thing. Start the loop running, then after a few seconds stop it with this method.
It is often necessary to "start fresh", and that is done by resetting the Notebook kernel with Kernel→Reset. After you've done this, you'll need to re-execute cells from the Cells menu or by selectively using SHIFT-ENTER in cells of interest.
In [ ]:
for i in xrange(1000):
print i
time.sleep(0.5)
IPython extends the functionality of the standard python interactive interpreter.
⁇ Should you ever use IPython to run Python programs?
⁇ What are some features of IPython?
IPython should only be used interactively, never ipython myprog.py.
Extended commands and syntax are only meant for interactive use, not to be added to Python modules.
Many great features:
cd ls cp etc.)
In [ ]:
pwd
In [ ]:
ls data
In [ ]:
# tab completion on attributes and functions with code-insight
time.
# NOTE: module must be loaded already
In [ ]:
# Help text
time?
In [ ]:
# tab completion on file names
fh = open('data/cla
In [ ]:
ls data/eur
In [ ]:
%%bash
for f in data/*.dat; do
echo -n $f " "
cat $f | wc -l
done
In [ ]:
%load_ext autosave
%autosave 300
In [ ]:
time?
In [ ]:
%run power
you can run ipython yourself:
Install it (if you don't have it already):
$ pip install ipythonRun it from the command line:
$ ipythonAnd if you have ipython then you can start your own ipython notebook:
$ ipython notebook --pylab=inlineBut for now, stick with Wakari
import math and execute it.math.<TAB>math.cos(<TAB>math.atan?%run weather.pyweather.py in the left menubar file browser to open it in a file editor. Take a look at what it is doing.data/europe-cities-temp.dat and prints out each line.Click on New Notebook
This will create a notebook with the name "Untitled X". Change the notebook name to "Weather Analysis" and save it.
Make the first cell Markdown and enter a title similar to the one used for this notebook, including a title and your name.
Add in a code cell with the autosave extension loaded and activated.
Add the following code block and execute it:
If you get an error regarding basemap then go to the Terminals tab in Wakari and start a Shell terminal then execute (don't inculde the dollar sign):
$ conda install basemap
In [ ]:
from mpl_toolkits.basemap import Basemap
import matplotlib.pyplot as plt
import numpy as np
fig = plt.figure(figsize=(12,12))
map = Basemap(projection='merc', lat_0 = 50, lon_0 = 5,
resolution = 'i',
llcrnrlon=-15, llcrnrlat=35,
urcrnrlon=20, urcrnrlat=62)
map.drawcoastlines()
map.drawcountries()
map.fillcontinents(color = 'coral')
map.drawmapboundary()
plt.show()
Share your notebook by clicking on the [Share] button beside it in the left menubar file manager. Provide a short description. Do not check "Share Environment" or set a password.
Try sharing your link with one of your tutorial buddies. Make sure you can open and view it in your browser.
Click on the code link near the top to hide the code in the notebook. Depending on the notebook content, this is sometimes a more useful and readable view.
This published notebook view is a static, read-only snapshot of your notebook. Others can clone your notebook into their own Wakari environment, or they can download the notebook to load in their own IPython Notebook environment.
Return to your "live" notebook and enter the code below after your import statements, but before the figure is generated into the existing code cell:
In [ ]:
'Reykjavik 61 (64,45,"N") (21,56,"W") 0.9 6.2 3.7 9.8 7.1 12.4 8.8 13.9'
regexp = "".join(['([^\t]+)\t([^\t]+)\t\((\d+),(\d+),"(\w)"\)\s+\((\d+),(\d+),',
'"(\w)"\)\t([^\t]+)\t([^\t]+)\t([^\t]+)\t([^\t]+)\t([^\t]+)\t([^\t]+)',
'\t([^\t]+)\t([^\t]+)\n'])
ndtype=[('city',str),
('elev', float),
('lat', float),
('lats', float),
('latns', '|S1'),
('lon', float),
('lons', float),
('lonew', '|S1'),
('apr_min', float),
('apr_max', float),
('may_min', float),
('may_max', float),
('jun_min', float),
('jun_max', float),
('jul_min', float),
('jul_max', float),
]
with open('data/europe-cities-temp-mod.dat') as fh:
cities = np.fromregex(fh, regexp, ndtype)
print cities.size
print cities['jul_max']
This should print out 17 and a list of values which are the July maximum temperatures for 17 European cities. If it does not, and you can't see what the problem is, ask your tutorial buddy for debugging help.
Finally, add the code below into the code cell just before the plt.show() statement.
In [ ]:
lat = cities['lat'] + cities['lats']/60
lon = cities['lon'] + cities['lons']/60
lon = lon*np.where(cities['lonew'] == 'E', 1, -1)
x,y = map(lon, lat)
map.plot(x, y, 'bo', markersize=30, alpha=0.5)
The last few steps are to save the figure as a file in PDF and PNG formats, save your work, and then share the notebook again. This time we will also share a bundle of all the work, for comparison.
In [ ]:
plt.savefig('weather.pdf')
plt.savefig('weather.png')
Save your Notebook.
Click the [Share] button in the Wakari file manager to re-publish your updated notebook. Check that the link works.
Click the Go Back link in the Wakari file manager to return to your home directory. Now click once on the wakari_tutorial directory to select it. You can now click the [Share] button in the menu bar above the file manager (middle button: box with an arrow going out of it). As before, do not share your environment.