Command mode vs Edit mode

By default we are in COMMAND mode

Press ENTER the edit the current cell

Press ESC to switch back to command mode

Main command mode shortcuts

Notebook control:

00 : Restart the notebook

Cells control:

Up/Down arrows : move up-down on cells
a : add cell above
b : add cell below
x : delete current cell

Editing cells:

Return : enter edit mode for current cell
Control+/ : Toggle code comment
Ctrl+Shift+- : Split cell at cursor position
Esc : return to command mode

Executing cells:

Shift+Return : execute the content of the current cell

More shortcuts listed under "Help" => "Keyboard shortcuts"

Cells editing

Cells have a type, which can be changed using shortcuts or the dedicated dropdown menu.
This is an example of text cell, where you can use Markdown tags to format your text.

You can also highlight chunks of code in almost any langauge

Example of Bash script:

#!/bin/bash
# A useless script
for i in $(seq 10); do
    echo  Hello World
done

Example of C fragment:

29 /*
  28  * System energy normalization
  27  * Returns the normalized value, in the range [0..SCHED_LOAD_SCALE],
  26  * corresponding to the specified energy variation.
  25  */
  24 static inline int
  23 normalize_energy(int energy_diff)
  22 {
  21         u32 normalized_nrg;
  20         int max_delta;
  19 
  18 #ifdef CONFIG_SCHED_DEBUG
  17         /* Check for boundaries */
  16         max_delta  = schedtune_target_nrg.max_power;
  15         max_delta -= schedtune_target_nrg.min_power;
  14         WARN_ON(abs(energy_diff) >= max_delta);
  13 #endif
  12 
  11         /* Do scaling using positive numbers to increase the range */
  10         normalized_nrg = (energy_diff < 0) ? -energy_diff : energy_diff;
   9 
   8         /* Scale by energy magnitude */
   7         normalized_nrg <<= SCHED_LOAD_SHIFT;
   6 
   5         /* Normalize on max energy for target platform */
   4         normalized_nrg = reciprocal_divide(
   3                         normalized_nrg, schedtune_target_nrg.rdiv);
   2 
   1         return (energy_diff < 0) ? -normalized_nrg : normalized_nrg;
5292 }

Code flow vs execution flow

Normally cells contains code, which is executed when Shift+Return is pressed



In [28]:

    
a = 1
b = 2



In [29]:

    
def my_simple_sum(a, b):
    """Simple addition
    :param a: fist number
    :param b: second number
    """
    print "Sum is:", a+b



In [33]:

    
my_simple_sum(a,b)









    



Sum is: 102



In [32]:

    
# Further down in the code we do some changes
a = 100
# than we can go back and re-execute just the previous cell

Access to documentation and Code completion



In [24]:

    
# Use TAB to complete the function name
# Use SHIFT+Tab after the '(' to access 
my_simple_sum(2,3)









    



Sum is: 5

Local shell commands execution

We can use a "!" at the beginning of a line to execute that command in a local shell



In [34]:

    
!pwd









    



/home/derkling/Code/lisa/ipynb/tutorial



In [36]:

    
!date









    



Fri Feb 26 17:52:47 CET 2016

We can also use variables as parameters by passing them wrapped in "{}"



In [57]:

    
folder = "../"
!ls -la {folder} | wc -l

Output of a local shell command can also be captured, for example to be post-processed in python



In [68]:

    
output = !find ../../ipynb/ -name "*.ipynb"

print "Available notebooks:"
for line in output:
    print line.replace('../../ipynb/', '   ')









    



Available notebooks:
   sched_tune/stune_juno_taskonly_rampL.ipynb
   sched_tune/stune_oak_rampL.ipynb
   sched_tune/stune_juno_rampL.ipynb
   devlib/examples/cgroups.ipynb
   profiling/kernel_functions_profiling.ipynb
   trappy/example_custom_events.ipynb
   sched_dvfs/smoke_test.ipynb
   wlgen/simple_rtapp.ipynb
   wlgen/rtapp_examples.ipynb
   tutorial/01_IPythonNotebooksUsage.ipynb
   tutorial/.ipynb_checkpoints/01_IPythonNotebooksUsage-checkpoint.ipynb
   utils/testenv_example.ipynb
   utils/executor_example.ipynb

How can be used?

Exporting to scripts

A notebook can be exported as a standalone script by via the "File" => "Download as" menu

Generate reports by exporting to HTML

By mixing code and markdown formatted comments it's quite easy to generate an HTML report which can be exported via the "File" => "Download As" menu. PDF format is also supported but it requires some configuration of the backend.

Useful links:

Complete tutorial on Notebooks: http://nbviewer.jupyter.org/github/ipython/ipython/blob/3.x/examples/Notebook/Index.ipynb
Markdown syntax: http://nbviewer.jupyter.org/github/ipython/ipython/blob/3.x/examples/Notebook/Working%20With%20Markdown%20Cells.ipynb