|
8 bits
|
1 byte
|
|
16 bits
|
word
|
|
32 bits
|
double word
|
|
always at least 1 byte
|
add one to first number in sequence to make negative, add zeros to get full byte
|
|
-program adjusts itself automatically to fit data, end result is a program trained to achieve a given task
|
|
Supervised learning-given examples of input and desired outputs, predict outputs on future unseen inputs(classification, regression, time series)
|
|
Unsupervised learning-creates a new representation of the input
|
|
Reinforcement learning-learning action to maximize payoff
|
|
Types of supervised learning tasks
|
|
1. classification- predict which predefined set of classes and example belongs to
|
|
2. regression - predict a real value
|
|
2. probability estimation - estimate probability of an event
|
|
Sensitivity = fraction of positive examples predicted to be positive
|
|
TP/(TP+FN)
|
|
Specificity= proportion of negative examples predicted negative
|
|
TN/(FP+TN)
|
|
False-positive rate(FPR)= negatives predicted to be positive
|
|
FP/(FP+TN)
|
|
Set
|
stores an unordered set of objects, CANT BE INDEXED, no duplicates, only immutable objects contained
|
|
myset = set([])
|
creates a set from a list
|
|
len(myset)
|
length of the set
|
|
if x in myset:
|
evaluates boolean condition
|
|
for element in myset:
|
iterate through set
|
|
for set1 = A and set2 = B, A&B = intersection
|
elements in common to A and B
|
|
A | B or A.union(B)
|
union of two sets
|
|
A-B
|
difference of sets, elements in A that are not in B
|
|
all these rules can be applied to multiple sets
|
|
range(start, end, step)
|
(start default is zero, step default 1)
|
|
range(5)
|
0 1 2 3 4
|
|
range(3,8)
|
3 4 5 6 7
|
|
range(len(seq))
|
sequence of index of values in seq
|
|
range(2,12,3)
|
2 5 8 11
|
|
def functionname( arguments):
|
defines a function of given name with given arguments
|
|
return
|
only returns a certain value or string generated by the function, doesn't print, exits at this value
|
|
list = [[0 for i in range(ncols)] for j in range(nrowns)]
|
creates a two dimensional list of nrows and ncols filled with zeros
|
|
lst = []
|
creates empty list
|
|
lst.append(val)
|
adds item to list at end
|
|
lst.extend(seq)
|
add sequence of items at end
|
|
lst.insert(idx, val)
|
inserts item at index
|
|
lst.remove(val)
|
remove first item in the list with value val
|
|
lst.pop(idx)
|
remove and return item at index
|
|
lst.sort()/lst.reverse()
|
sort/reverse list in place
|
|
lst.min()/.max()
|
finds the min/max
|
|
.plot(x,y, color)
|
plots the x values to x values in a certain color
|
|
.show()
|
shows the graph
|
|
.ylabel(name)
|
names y axis
|
|
.xlabel(name)
|
names x axis
|
|
.savefig( figname)
|
saves image under figname
|
|
class
|
defines attributes(information we want to keep together) and methods( operations want to be able to perform on that data)
|
|
class some_class_name:
|
defines a class of some_class_name
|
|
def __int__(self, other attributes):
|
defines and initializes attributes
|
|
object.method()
|
calls method on an object of that class
|
|
Sytax errors
|
violation of a writing rule
|
|
Exceptions(runtime)
|
syntax is fine, some other thing is occasionally flawed:ZeroDivisionError(can't divide by zero), NameError(can't access the variable), IndexError(When the index you are attempting to access does not exist), TypeError( operation on non-compatible type)
|
|
Logical (semantic) errors
|
code runs, but doesn't do what its supposed to
|
|
try:
|
does something that may cause an exception
|
|
except <SomeExceptionType>:
|
do something to handle the exception
|
|
raise[ exception object ]
|
raises a certain, defined type of exception
|
|
try, except, else
|
try something, that could cause the exception, if its fine, go to else
|
|
finally:
|
adds statement that happens no matter what
|
|
for a, b in zip(list 1, list 2):
|
iterates over elements of two lists in parallel, yields a tuple with both iterations
|
|
f = open(myfile, 'x')
|
opens the file, x=r for reading only, x=w for writing only, x=a for appending, x=b for file in binary format, x=wr+ reading and writing and so on
|
|
.read(size)
|
reads the entire file, returns a single string, size is optional number of characters
|
|
.readlines(size)
|
reads all lines and returns them as a list of strings
|
|
.rstrip()
|
returns a string with the end-of-line character(\n) removed from the end of the string
|
|
.readline()
|
reads single line from file, returns empty string if end of file
|
|
.close()
|
closes file
|
|
gzip.open()
|
interface to read/write compressed files, .gz extension
|
|
dict = {}
|
sets up an empty dictionary
|
|
dict[key] = value
|
sets value of said at that key to value
|
|
dict[key]
|
calls the value at that key
|
|
del dict[key]
|
deletes a key from dict
|
|
dict.clear()
|
clears the dict of all keys
|
|
dict.update(dict2)
|
can update/add associations from another
|
|
dict.keys()/dict.items()/dict.values()
|
looks at either all keys, values, or all items in the dict
|
|
dict.pop(key)
|
removes element associated to that key
|
|
dict.popitem()
|
removes key, value pair and returns it as a tuple
|
|
dict.get(key)
|
returns value at that key
|
|
dict.setdefault(key)
|
either returns value of the key or creates the given key if not previously existent
|
No comments yet.
Cheatography n. The art of mapping complex subjects into clear, instantly navigable reference guides.
