Notebook for the Python lesson at Weill Cornell Medical College¶
(Time 3 hours 15 minutes)
Python jargon¶
- "Python" refers to the programming language itself. You use Python to write source code and give instructions to your computer.
- Python has two dialects: Python 2 (we're using that) and Python 3 (newer, but not sticking yet)
- Python has only a few words (keywords), like 20~ish.
printis a keyword (in Python 2, not in 3)
- "python" refers to the interpreter, a program that translates source code into actual instructions to the machine
- you call the interpreter typing
pythonin your terminal - working in the interpreter can be hard (ugly and error-prone)
- you call the interpreter typing
- "ipython" is a wrapper around the interpreter, which adds some nice features, but you still have to work in a command line interface
- you call the "ipython" typing
ipythonin your terminal
- you call the "ipython" typing
- "ipython notebook" is another "wrapper" that puts everything into your browser
- you call the "ipython notebook" (now called "jupyter") typing
ipython notebookin your terminal
- you call the "ipython notebook" (now called "jupyter") typing
A simple command (actually is called a statement) in Python¶
In [1]:
print "Hello"
Names and objects¶
In [2]:
weight_kg = 55
In [3]:
weight_kg
Out[3]:
In [4]:
print weight_kg
In [5]:
print "Weight in pounds is ", 2.2 * weight_kg
In [6]:
weight_lb = 2.2 * weight_kg
In [7]:
print weight_lb
In [8]:
weight_kg = 100
In [9]:
print weight_kg
In [10]:
print weight_lb
Loading a library of functions (is called a module) from your hard drive into this program¶
Numpy is a library for some math in Python
In [11]:
import numpy
We creating a new name (data) to store the result of calling a function in the numpy module that reads text files (numpy.loadtxt)
In [14]:
data = numpy.loadtxt(fname="data/inflammation-01.csv", delimiter=',')
In [15]:
print data
type is a Python built-in function (is loaded by default when you start any interpreter) which tells you the kind of object referred by a name.
Object is the generic concept for "things" in Python, everything is an object. Objects represent how the "thing" is stored in the computer memory, for example they have instructions on how your computer reads the object from the memory, so you may get an error if you are trying to process an object as the wrong kind. More importantly for you, objects have functions associated to them (called methods). Each type of object may have its own methods.
In [16]:
print type(data)
In [17]:
print type(weight_lb)
In [18]:
data.shape
Out[18]:
In [20]:
weight_lb.shape
Indexing and slicing¶
In [22]:
data[3:4, 20]
Out[22]:
In [22]:
data[:4, 30:]
Out[22]:
In [22]:
data[30, 20]
Out[22]:
In [28]:
sample = data[4:8, 1:3]
In [29]:
print sample
In [ ]:
### Operators and methods for Numpy arrays
In [30]:
double_sample = 2 * sample
In [31]:
print double_sample
In [32]:
data.mean()
Out[32]:
In [33]:
data.max()
Out[33]:
In [34]:
data.min()
Out[34]:
Pause and reflect on the amount of work that your computer has done when you called any of this methods. This is called high-level programming language and means you don't need to know much about how your computer does its thing.
Slicing a numpy array gives you another numpy array (nice, we still have all methods available)
In [35]:
patient_0 = data[0, :]
In [36]:
print patient_0
In [37]:
type(patient_0)
Out[37]:
In [38]:
patient_0.shape
Out[38]:
In [39]:
patient_0.mean()
Out[39]:
In [41]:
data.mean(axis=1)
Out[41]:
Plotting¶
You can import parts of a module using from, in this case the pyplot part of matplotlib.
In [ ]:
from matplotlib import pyplot
In [45]:
%matplotlib inline
image = pyplot.imshow(data)
pyplot.show(image)
In [46]:
ave_inflammation = data.mean(axis=0)
ave_plot = pyplot.plot(ave_inflammation)
pyplot.show(ave_plot)
In [47]:
max_plot = pyplot.plot(data.max(axis=0))
pyplot.show(max_plot)
Putting all together in one nice piece of code¶
In [48]:
import numpy as np
from matplotlib import pyplot as plt
data = np.loadtxt(fname="data/inflammation-01.csv", delimiter=",")
fig = plt.figure(figsize=(10.0, 3.0))
axes1 = fig.add_subplot(1, 3, 1)
axes2 = fig.add_subplot(1, 3, 2)
axes3 = fig.add_subplot(1, 3, 3)
axes1.set_ylabel('average')
axes1.plot(data.mean(axis=0))
axes2.set_ylabel('max')
axes2.plot(data.max(axis=0))
axes3.set_ylabel('min')
axes3.plot(data.min(axis=0))
fig.tight_layout()
plt.show(fig)
For loops¶
In [51]:
word = "lead"
In [52]:
word
Out[52]:
In [55]:
print word[0]
print word[1]
print word[2]
print word[3]
In [58]:
length = 0
for char in word:
length = length + 1
print char
print "The number of letters ", length
In [59]:
print char
In [60]:
print length
In [61]:
len(word)
Out[61]:
In [64]:
range(2, 8, 2)
Out[64]:
In [66]:
for number in range(1,4):
print number
In [71]:
result = 1
for times in range(3):
result *= 5
print result
Lists¶
In [73]:
l = range(5)
In [74]:
type(l)
Out[74]:
In [75]:
odd = [1, 3, 5]
In [76]:
type(odd)
Out[76]:
In [77]:
print odd
In [81]:
odd[1:]
Out[81]:
In [83]:
odd[-2]
Out[83]:
In [84]:
names = ["Newton", "Darwin", "Tuinrg"]
In [85]:
names[-2]
Out[85]:
In [86]:
darwin = names[-2]
In [87]:
print darwin
In [88]:
names[-1] = "Turing"
In [89]:
names
Out[89]:
In [90]:
darwin
Out[90]:
In [91]:
darwin[0]
Out[91]:
In [92]:
darwin[0] = 'd'
In [93]:
odd
Out[93]:
In [94]:
odd.append(7)
In [95]:
odd
Out[95]:
In [96]:
odd.reverse()
In [97]:
odd
Out[97]:
In [98]:
odd.append('john')
Lists can have different types of objects
In [99]:
odd
Out[99]:
In [101]:
type(odd[0])
Out[101]:
In [102]:
type(odd[-1])
Out[102]:
Using glob to get a list of files in our filesystem¶
Sorta ls in the shell, but storing the result into a Python list (nice!)
In [103]:
import glob
In [105]:
print glob.glob('*.ipynb')
In [106]:
print glob.glob('./data/*.csv')
In [107]:
filenames = glob.glob("data/*csv")
In [108]:
print filenames
Putting all together: loop and analyze a bunch of files¶
In [111]:
for f in filenames[:3]:
print "Analyzing file: ", f
analyze_inflammation(f)
In [153]:
def analyze(filename):
"""Analyze inflammation data and make plots
"""
data = np.loadtxt(fname=filename, delimiter=",")
fig = plt.figure(figsize=(10.0, 3.0))
axes1 = fig.add_subplot(1, 3, 1)
axes2 = fig.add_subplot(1, 3, 2)
axes3 = fig.add_subplot(1, 3, 3)
axes1.set_ylabel('average')
axes1.plot(data.mean(axis=0))
axes2.set_ylabel('max')
axes2.plot(data.max(axis=0))
axes3.set_ylabel('min')
axes3.plot(data.min(axis=0))
fig.tight_layout()
plt.show(fig)
Conditionals¶
In [116]:
weight_kg = 20
In [117]:
if weight_kg > 80:
print "You're over 80 kg"
else:
print "You're under 80 kg"
In [129]:
number = -3
In [132]:
if (number >= 0):
print "Positive"
else:
print "Negative"
Functions¶
In [135]:
(80 - 32) * 5/9 + 273.15
Out[135]:
In [136]:
def fahr_to_kelvin(temp):
return (temp - 32) * 5.0/9 + 273.15
In [150]:
def kelvin_to_celsius(temp):
'''Converts temperatures in Kelvin to Celsius
More stuff
'''
result = temp - 273.15
return result
In [ ]:
def farh_to_celsius(temp):
kelvin = fahr_to_kelvin(temp)
result = kelvin_to_celsius(kelvin)
return result
In [152]:
kelvin_to_celsius(fahr_to_kelvin(0))
Out[152]:
In [138]:
fahr_to_kelvin(80)
Out[138]:
In [139]:
temp_in_kelvin = fahr_to_kelvin(100)
In [140]:
temp_in_kelvin
Out[140]:
Be careful with dividing numbers
In [3]:
10/9
Out[3]:
In [143]:
10.0/9
Out[143]:
In [154]:
help(analyze)
In [155]:
help(np.loadtxt)