1. Pierre Denis Avatar Pierre Denis
  2. Python projects
  3. Lea

Wiki

Clone wiki

Lea / LeappSnippets

View History

Table of Content

Creating a probability distribution

lea> die1 = ?(1,2,3,4,5,6)
lea> die1
1 : 1/6
2 : 1/6
3 : 1/6
4 : 1/6
5 : 1/6
6 : 1/6

or, equivalently,

lea> die1 = ?([1,2,3,4,5,6])
lea> die1 = ?(range(1,7))
lea> die1 = ?(range(6)) + 1
lea> die1 = 1 + ?(range(6))
lea> die1 = ?(6-i for i in range(6))
lea> valuesTuple = (1,2,3,4,5,6)
lea> die1 = ?(valuesTuple)
lea> valuesIter = (6-i for i in range(6))
lea> die1 = ?(valuesIter)
lea> die1 = ?{1: 1, 2: 1, 3: 1, 4: 1, 5: 1, 6: 1}
lea> die1 = ?{1: 1/6, 2: 1/6, 3: 1/6, 4: 1/6, 5: 1/6, 6: 1/6}

Examples with strings: 

lea> fairWeather = ?('sunny','cloudy','raining')
lea> fairWeather
 cloudy : 1/3
raining : 1/3
  sunny : 1/3
lea> weather = ?{'sunny': 68%, 'cloudy': 7%, 'raining': 25%}
lea> weather
 cloudy :  7/100
raining : 25/100
  sunny : 68/100

Since Python strings are iterable sequences, a letter frequency counter is made easy:

lea> ?('ALEA JACTA EST')
  : 2/14
A : 4/14
C : 1/14
E : 2/14
J : 1/14
L : 1/14
S : 1/14
T : 2/14

Note: to define a "singleton" distribution (certain value), append a coma (standard Python trick):

lea> ?('ALEA JACTA EST',)
ALEA JACTA EST : 1

Boolean random variables:

lea> flip = ?:(1/2)
lea> flip
False : 1/2
 True : 1/2

equivalent to

lea> flip = ?:(50%)
lea> flip = ?:(0.5)
lea> flip = ?(False,True)
lea> flip = ?{False: 1/2, True: 1/2}
lea> flip = ?{False: 50%, True: 50%}
lea> flip = ?{False: 0.5, True: 0.5}

and

lea> bflip = ?:(1/4)
lea> bflip
False : 3/4
 True : 1/4

equivalent to

lea> bflip = ?:(25%)
lea> bflip = ?:(0.25)
lea> bflip = ?{False: 3/4, True: 1/4}
...

Cloning a distribution

(in order to have independent random variable)

lea> die2 = ?die1

Adding distributions

lea> dice = die1 + die2
lea> dice
 2 : 1/36
 3 : 2/36
 4 : 3/36
 5 : 4/36
 6 : 5/36
 7 : 6/36
 8 : 5/36
 9 : 4/36
10 : 3/36
11 : 2/36
12 : 1/36

or, equivalently (well, if no conditional probability involved - see below!),

lea> dice = die1 + ?die1
lea> dice = sum((die1,die2))
lea> dice = sum(?die1 for i in range(2))
lea> dice = ?[2]die1

Displaying probabilities

lea> :. dice
 2 : 0.027778
 3 : 0.055556
 4 : 0.083333
 5 : 0.111111
 6 : 0.138889
 7 : 0.166667
 8 : 0.138889
 9 : 0.111111
10 : 0.083333
11 : 0.055556
12 : 0.027778
lea> :% dice
 2 :   2.8 %
 3 :   5.6 %
 4 :   8.3 %
 5 :  11.1 %
 6 :  13.9 %
 7 :  16.7 %
 8 :  13.9 %
 9 :  11.1 %
10 :   8.3 %
11 :   5.6 %
12 :   2.8 %
lea> :% bflip
False :  75.0 %
 True :  25.0 %

Extracting individual probabilities

lea> dice@7
1/6
lea> :. dice@7
0.16666666666666666
lea> :% dice@7
16.666667 %
lea> (dice <= 6) @True
5/12
lea> :% (dice <= 6) @True
41.666667 %
lea> bflip@True
1/4
lea> :. bflip@True
0.25

Note that @True can be shortcut as @:

lea> :% (dice <= 6) @
41.666667 %
lea> :. bflip@
0.25

Evaluating conditions

lea> die1 == 2
False : 5/6
 True : 1/6
lea> die1 <= 2
False : 2/3
 True : 1/3
lea> :% die1 <= 2
False :  66.7 %
 True :  33.3 %
lea> (2 <= die1) & (die1 < 4)
False : 2/3
 True : 1/3
lea> ~((2 <= die1) & (die1 < 4))
False : 1/3
 True : 2/3
lea> (2 > die1) | (die1 >= 4)
False : 1/3
 True : 2/3
lea> (2 <= die1) & (die2 < 4)
False : 7/12
 True : 5/12
lea> (2 <= die1) & (dice < 4)
False : 35/36
 True :  1/36
lea> ?[50]die1 == 60
False : 808281277464764060643139600393722578321/808281277464764060643139600456536293376
 True :                             62813715055/808281277464764060643139600456536293376
lea> :. ?[50]die1 == 60
False : 1.000000
 True : 0.000000

Calculating probability distribution indicators

lea> dice.mean
7.0
lea> dice.stdev
2.41522945769824
lea> dice.entropy
3.2744019192887714

Generating random samples

lea> dice $
7
lea> dice $
5
lea> dice $(20)
(5, 4, 7, 10, 6, 9, 8, 6, 10, 4, 7, 6, 7, 6, 3, 10, 7, 5, 9, 6)
lea> flip $(20)
(False, True, True, False, True, True, False, False, False, True, False, True, True, True, False, True, False, False, True, True)
lea> bflip $(20)
(False, False, True, False, False, False, False, False, False, True, False, False, True, False, False, True, False, False, True, True)

Note that dice $(20) generates something equivalent to the the results of any of the following expressions 

lea> tuple(dice$ for i in range(20))
lea> tuple(die1$ + die2$ for i in range(20))
... but faster!

Controlling the frequencies in a random sample: 

lea> ?(dice $(36000))
 2 :  931/36000
 3 : 2000/36000
 4 : 3065/36000
 5 : 3859/36000
 6 : 5088/36000
 7 : 6010/36000
 8 : 5010/36000
 9 : 3964/36000
10 : 3030/36000
11 : 2047/36000
12 :  996/36000
lea> :% ?(dice $(36000))
 2 :   2.7 %
 3 :   5.5 %
 4 :   8.6 %
 5 :  10.6 %
 6 :  13.6 %
 7 :  17.0 %
 8 :  13.9 %
 9 :  11.3 %
10 :   8.2 %
11 :   5.5 %
12 :   3.0 %

The return of Caesar, the latin parrot... 

lea> latinParrot = ?('ALEA JACTA EST')
lea> ''.join(latinParrot $(30))
'TA   AECETAAATA AE ETJ AASLAAE'

Calculating a cartesian product

lea> ?*(die1,die2)
(1, 1) : 1/36
(1, 2) : 1/36
(1, 3) : 1/36
(1, 4) : 1/36
(1, 5) : 1/36
(1, 6) : 1/36
(2, 1) : 1/36
...
(6, 5) : 1/36
(6, 6) : 1/36
lea> ?*(die1,die2,die1+die2)
 (1, 1, 2) : 1/36
 (1, 2, 3) : 1/36
 (1, 3, 4) : 1/36
...
(6, 5, 11) : 1/36
(6, 6, 12) : 1/36

Applying a function

Putting a ? in front of a function applied on Lea argument(s) builds a new Lea distribution, by applying the function on each inner values:

lea> 'die1 = ' + ?str(die1)
die1 = 1 : 1/6
die1 = 2 : 1/6
die1 = 3 : 1/6
die1 = 4 : 1/6
die1 = 5 : 1/6
die1 = 6 : 1/6

Note the difference if the ? is omitted: 

lea> 'die1 = ' + str(die1)
'die1 = 1 : 1/6\n2 : 1/6\n3 : 1/6\n4 : 1/6\n5 : 1/6\n6 : 1/6'
Without the ? prefix, the str function is applied on the whole distribution, instead of individual values : the result is a standard Python string, not a Lea distribution.

User-defined functions

lea> def f(x,y):
 ...     return '%02d!' % (x+y)
lea> ?f(die1,die2)
02! : 1/36
03! : 2/36
04! : 3/36
05! : 4/36
06! : 5/36
07! : 6/36
08! : 5/36
09! : 4/36
10! : 3/36
11! : 2/36
12! : 1/36

Applying an instance method

lea> die1.to_bytes(2,"big")
b'\x00\x01' : 1/6
b'\x00\x02' : 1/6
b'\x00\x03' : 1/6
b'\x00\x04' : 1/6
b'\x00\x05' : 1/6
b'\x00\x06' : 1/6

Getting an instance attribute

lea> complexDice = die1 + die2*1j
lea> complexDice
(5+2j) : 1/36
(1+1j) : 1/36
(2+1j) : 1/36
...
(6+3j) : 1/36
(6+5j) : 1/36
lea> complexDice.imag
1.0 : 1/6
2.0 : 1/6
3.0 : 1/6
4.0 : 1/6
5.0 : 1/6
6.0 : 1/6

Evaluating conditional probabilities

What follows the exclamation mark is a given fact, expressed as a condition expression. The expression before the exclamation mark is a distribution that is evaluated assuming that the given condition is certainly true.

lea> dice = die1 + die2
lea> dice ! dice < 5
2 : 1/6
3 : 2/6
4 : 3/6
lea> dice ! die1+die2 < 5
2 : 1/6
3 : 2/6
4 : 3/6
lea> dice ! (4 <= dice) & (dice < 7)
4 : 3/12
5 : 4/12
6 : 5/12
lea> dice ! (die1 <= 3) & (die2 <= 3)
2 : 1/9
3 : 2/9
4 : 3/9
5 : 2/9
6 : 1/9
lea> (5 <= dice) ! (die1 <= 3) & (die2 <= 3)
False : 2/3
 True : 1/3
lea> ?*(die1,die2) ! dice == 5
(1, 4) : 1/4
(2, 3) : 1/4
(3, 2) : 1/4
(4, 1) : 1/4
lea> ?*(die1,die2,dice) ! dice <= 5
(1, 1, 2) : 1/10
(1, 2, 3) : 1/10
(1, 3, 4) : 1/10
(1, 4, 5) : 1/10
(2, 1, 3) : 1/10
(2, 2, 4) : 1/10
(2, 3, 5) : 1/10
(3, 1, 4) : 1/10
(3, 2, 5) : 1/10
(4, 1, 5) : 1/10

Indexing / slicing

Provided that the values in the distribution are indexable (e.g. strings, tuples), indexing made on a Lea instance is propagated in the inner values, obeying Python index numbering:

lea> name = ?('Gino', 'Guy', 'Jack', 'Lea', 'Leon', 'Loth', 'Lucienne', 'Pierre', 'Piotr', 'Rachel')
lea> name[0]
G : 2/10
J : 1/10
L : 4/10
P : 2/10
R : 1/10
lea> name[1]
a : 2/10
e : 2/10
i : 3/10
o : 1/10
u : 2/10
lea> name[-1]
a : 1/10
e : 2/10
h : 1/10
k : 1/10
l : 1/10
n : 1/10
o : 1/10
r : 1/10
y : 1/10
lea> name[0] + name[-1]
Go : 1/10
Gy : 1/10
Jk : 1/10
La : 1/10
Le : 1/10
Lh : 1/10
Ln : 1/10
Pe : 1/10
Pr : 1/10
Rl : 1/10

Slicing is also feasible:

lea> name[1:3]
ac : 2/10
ea : 1/10
eo : 1/10
ie : 1/10
in : 1/10
io : 1/10
ot : 1/10
uc : 1/10
uy : 1/10
lea> 'T'+name[1:] ! name[0]=='L'
     Tea : 1/4
    Teon : 1/4
    Toth : 1/4
Tucienne : 1/4

Misc

Playing cards

lea> cards = ?('A23456789TJQK') + ?('♥♣♦♠')
lea> cards
2♠ : 1/52
2♣ : 1/52
2♥ : 1/52
2♦ : 1/52
3♠ : 1/52
3♣ : 1/52
3♥ : 1/52
3♦ : 1/52
4♠ : 1/52
4♣ : 1/52
4♥ : 1/52
4♦ : 1/52
5♠ : 1/52
5♣ : 1/52
5♥ : 1/52
5♦ : 1/52
6♠ : 1/52
6♣ : 1/52
6♥ : 1/52
6♦ : 1/52
7♠ : 1/52
7♣ : 1/52
7♥ : 1/52
7♦ : 1/52
8♠ : 1/52
8♣ : 1/52
8♥ : 1/52
8♦ : 1/52
9♠ : 1/52
9♣ : 1/52
9♥ : 1/52
9♦ : 1/52
A♠ : 1/52
A♣ : 1/52
A♥ : 1/52
A♦ : 1/52
J♠ : 1/52
J♣ : 1/52
J♥ : 1/52
J♦ : 1/52
K♠ : 1/52
K♣ : 1/52
K♥ : 1/52
K♦ : 1/52
Q♠ : 1/52
Q♣ : 1/52
Q♥ : 1/52
Q♦ : 1/52
T♠ : 1/52
T♣ : 1/52
T♥ : 1/52
T♦ : 1/52
lea> cards[0]
2 : 1/13
3 : 1/13
4 : 1/13
5 : 1/13
6 : 1/13
7 : 1/13
8 : 1/13
9 : 1/13
A : 1/13
J : 1/13
K : 1/13
Q : 1/13
T : 1/13
lea> cards[1]
♠ : 1/4
♣ : 1/4
♥ : 1/4
♦ : 1/4
lea> cards $
'8♣'
lea> cards $_()
('K♣', '4♠', '7♥', 'T♥', 'A♥', '6♣', '6♦', '9♠', '5♣', 'Q♥', 'T♣', 'K♠', '4♥', 'T♦', '8♣', 'A♦', 'J♣', 'A♠', 'J♥', '3♠', '4♦', 'K♦', '8♥', '7♣', '8♦', '6♥', 'T♠', '5♥', '3♥', '9♦', 'J♦', '3♣', '2♥', '9♣', '7♠', 'Q♣', '5♦', '4♣', 'A♣', '9♥', '2♦', '6♠', 'K♥', '2♠', 'Q♦', '5♠', '7♦', 'J♠', 'Q♠', '8♠', '2♣', '3♦')
lea> cards $_(13)
('9♠', '4♦', 'K♠', '9♦', '6♦', 'A♦', '3♥', '8♣', 'T♥', '5♦', '7♦', '9♣', 'Q♥')
lea> ' '.join(cards $_(13))
'8♦ 7♦ 5♦ K♦ 6♠ K♠ T♥ 2♦ 3♠ 9♣ 7♥ 2♠ T♣'
lea> ?(cards$(52000))
2♠ : 1018/52000
2♣ :  993/52000
2♥ : 1001/52000
2♦ :  974/52000
3♠ :  952/52000
3♣ : 1047/52000
3♥ : 1056/52000
3♦ :  978/52000
4♠ : 1057/52000
4♣ :  995/52000
4♥ :  941/52000
4♦ : 1052/52000
5♠ :  972/52000
5♣ :  971/52000
5♥ :  954/52000
5♦ :  985/52000
6♠ : 1043/52000
6♣ :  979/52000
6♥ :  972/52000
6♦ :  970/52000
7♠ :  996/52000
7♣ : 1002/52000
7♥ : 1052/52000
7♦ : 1031/52000
8♠ :  991/52000
8♣ :  995/52000
8♥ : 1016/52000
8♦ : 1016/52000
9♠ :  990/52000
9♣ :  959/52000
9♥ : 1015/52000
9♦ :  989/52000
A♠ : 1030/52000
A♣ : 1004/52000
A♥ :  993/52000
A♦ : 1019/52000
J♠ : 1019/52000
J♣ :  963/52000
J♥ : 1024/52000
J♦ : 1046/52000
K♠ :  953/52000
K♣ :  984/52000
K♥ :  994/52000
K♦ : 1075/52000
Q♠ : 1042/52000
Q♣ :  967/52000
Q♥ :  935/52000
Q♦ :  977/52000
T♠ : 1029/52000
T♣ :  994/52000
T♥ : 1004/52000
T♦ :  986/52000

Nice dice...

lea> diePic = ?('⚀⚁⚂⚃⚄⚅')
lea> diePic 
⚀ : 1/6
⚁ : 1/6
⚂ : 1/6
⚃ : 1/6
⚄ : 1/6
⚅ : 1/6
lea> ?[2]diePic 
⚀⚀ : 1/36
⚀⚁ : 1/36
⚀⚂ : 1/36
⚀⚃ : 1/36
⚀⚄ : 1/36
⚀⚅ : 1/36
⚁⚀ : 1/36
⚁⚁ : 1/36
⚁⚂ : 1/36
⚁⚃ : 1/36
⚁⚄ : 1/36
⚁⚅ : 1/36
⚂⚀ : 1/36
⚂⚁ : 1/36
⚂⚂ : 1/36
⚂⚃ : 1/36
⚂⚄ : 1/36
⚂⚅ : 1/36
⚃⚀ : 1/36
⚃⚁ : 1/36
⚃⚂ : 1/36
⚃⚃ : 1/36
⚃⚄ : 1/36
⚃⚅ : 1/36
⚄⚀ : 1/36
⚄⚁ : 1/36
⚄⚂ : 1/36
⚄⚃ : 1/36
⚄⚄ : 1/36
⚄⚅ : 1/36
⚅⚀ : 1/36
⚅⚁ : 1/36
⚅⚂ : 1/36
⚅⚃ : 1/36
⚅⚄ : 1/36
⚅⚅ : 1/36
lea> ''.join(diePic?(60))
'⚄⚃⚃⚅⚁⚂⚁⚁⚂⚅⚀⚃⚁⚂⚄⚀⚀⚃⚃⚁⚃⚀⚂⚃⚄⚁⚅⚄⚂⚄⚀⚄⚃⚀⚃⚂⚂⚀⚁⚀⚃⚃⚅⚄⚂⚅⚃⚅⚅⚀⚁⚅⚂⚁⚄⚁⚄⚃⚀⚂'

Bit counter

variant 1 : from integer to string

lea> flip1 = ?(0,1)
lea> flip2 = ?flip1
lea> flip1 + flip2
0 : 1/4
1 : 2/4
2 : 1/4
lea> 'flip1 = ' + ?str(flip1) + ', flip2 = ' + ?str(flip2) + ' => flip1+flip2 = ' + ?str(flip1+flip2)
flip1 = 0, flip2 = 0 => flip1+flip2 = 0 : 1/4
flip1 = 0, flip2 = 1 => flip1+flip2 = 1 : 1/4
flip1 = 1, flip2 = 0 => flip1+flip2 = 1 : 1/4
flip1 = 1, flip2 = 1 => flip1+flip2 = 2 : 1/4

variant 2 : from string to integer

lea> c = ?('01')
lea> c4 = ?[4]c
lea> c4
0000 : 1/16
0001 : 1/16
0010 : 1/16
0011 : 1/16
0100 : 1/16
0101 : 1/16
0110 : 1/16
0111 : 1/16
1000 : 1/16
1001 : 1/16
1010 : 1/16
1011 : 1/16
1100 : 1/16
1101 : 1/16
1110 : 1/16
1111 : 1/16
lea> f = lambda s: sum(int(d) for d in s)
lea> count = ?f(c4)
lea> count
0 : 1/16
1 : 4/16
2 : 6/16
3 : 4/16
4 : 1/16
lea> ?*(c4,count)
('0000', 0) : 1/16
('0001', 1) : 1/16
('0010', 1) : 1/16
('0011', 2) : 1/16
('0100', 1) : 1/16
('0101', 2) : 1/16
('0110', 2) : 1/16
('0111', 3) : 1/16
('1000', 1) : 1/16
('1001', 2) : 1/16
('1010', 2) : 1/16
('1011', 3) : 1/16
('1100', 2) : 1/16
('1101', 3) : 1/16
('1110', 3) : 1/16
('1111', 4) : 1/16
lea> c4 + ' -> ' + ?str(count)
0000 -> 0 : 1/16
0001 -> 1 : 1/16
0010 -> 1 : 1/16
0011 -> 2 : 1/16
0100 -> 1 : 1/16
0101 -> 2 : 1/16
0110 -> 2 : 1/16
0111 -> 3 : 1/16
1000 -> 1 : 1/16
1001 -> 2 : 1/16
1010 -> 2 : 1/16
1011 -> 3 : 1/16
1100 -> 2 : 1/16
1101 -> 3 : 1/16
1110 -> 3 : 1/16
1111 -> 4 : 1/16

Binary choices

lea> sendEMailToDREarlyNovember = ?:(99.9%)
lea> :% sendEMailToDREarlyNovember
False :   0.1 %
 True :  99.9 %
lea> 'I will ' + ?{'':99.9% ,'not ':0.1%} + 'send an e-mail to D. R. early November.'
I will not send an e-mail to D. R. early November. :   1/1000
    I will send an e-mail to D. R. early November. : 999/1000

Updated