Python - Generating a New List Randomly From Two Other Lists

Question

I have two lists, high and low bounds, that I would like to use as bounds to create a new list randomly between those values.

LOW     HIGH
19.35   45.04
18.09   44.05
17.44   42.07
16.46   41.95
18.38   43.45
20.78   46.04
24.99   49.12
25.22   50.94
24.67   49.32

I am trying to create a new list that has the length of list but has random values between each LOW and HIGH pair.

A new list might look like:

NEW LIST
38.26
33.70
18.47
41.35
36.77
21.95
25.23
29.18
49.32

Any help is appreciated, thanks!

So far this is what I have but it doesn't step down a row to update the lower and upper bounds:

Low = pd.Series(data['LOW']).values
High = pd.Series(data['HIGH']).values

newlist = []

for i in range(0, 8):
        newlist = random.uniform(Low, High)

Are these plain Python lists, or are they Pandas Series or something? — AMC, Nov 12 '19 at 23:07
Updated to show initialization of the Low and High, they are not dataframes, just series with no index. — bathtub2007, Nov 12 '19 at 23:11
Do you want the result as a Series? If I may ask, why do you have some poor orphan Series, all on their own? — AMC, Nov 12 '19 at 23:12
Let us [continue this discussion in chat](https://chat.stackoverflow.com/rooms/202236/discussion-between-alexander-cecile-and-bathtub2007). — AMC, Nov 12 '19 at 23:12
Possible duplicate of [Python List random choice](https://stackoverflow.com/questions/44355406/python-list-random-choice) — oreopot, Nov 15 '19 at 05:10

score 1 · Accepted Answer · edited Jun 20 '20 at 09:12

All hail the power of broadcasting!

import pandas as pd
import random
import numpy as np

# example data
lows = pd.Series(np.random.uniform(0, 50, 10))
highs = pd.Series(np.random.uniform(0, 50, 10))

nums_arr = random.uniform(low=lows, high=highs)

I especially like this solution because it directly creates a Series, no conversions necessary.

A few comments on your code:

1)

Variable names should generally follow the lower_case_with_underscores style. See PEP 8 for more on Python style.

2)

Low = pd.Series(data['LOW']).values
High = pd.Series(data['HIGH']).values

From the rest of the code you showed me, you already know that data is a DataFrame. Getting a column from a DataFrame (e.x: data['LOW']) returns a Series, so the call to pd.Series() is redundant. It also isn't clear why you then call .values().

Bear in mind that using Series.values is discouraged. From the docs:

Warning: We recommend using Series.array or Series.to_numpy(), depending on whether you need a reference to the underlying data or a NumPy array.

3)

newlist = []

for i in range(0, 8):
        newlist = random.uniform(Low, High)

I'm really struggling to find the words to explain this one. random.uniform is a function which takes two numbers, and produces a random float between the first two. In this case, however, random.uniform is operating on two Series, thanks to the magic of broadcasting. This is why despite not looking like it makes much sense, the code doesn't actually produce an error or crash.

Notice also that the code above doesn't actually do anything with the result in newlist, nor does newList depend at all on the variable i.

My guess is that you had something like this in mind:

# notice the variable name
new_nums_list = []

for i in range(low.size):
    curr_rand_num = random.uniform(low[i], high[i])
    new_nums_list.append(curr_rand_num)

Which could have also been written:

new_nums_list = [random.uniform(low[i], high[i]) for i in range(low.size)]

Thanks to the power of zip(), we could do even better:

new_nums_list = [random.uniform(curr_low, curr_high) for curr_low, curr_high in zip(low, high)]

Any further questions? :)

This is beyond helpful, appreciate both working code but where I might think about improving my own code. Thank you! — bathtub2007, Nov 13 '19 at 16:09

Konstantinos Katsantonis · Answer 2 · 2019-11-13T21:35:58.977

0

import random 
HIGH = [45.04, 44.05, 42.07, 41.95, 43.45, 46.04, 49.12, 50.94, 49.32] 
LOW = [19.35, 18.09, 17.44, 16.46, 18.38, 20.78, 24.99, 25.22, 24.67]  
result = []

for low,high in zip(LOW,HIGH): 
    result.append(random.uniform(low,high))

print(result)

Example output:

[20.23415150413731, 31.21737764031939, 33.39330466412157, 29.23878916462911, 33.23211876295469, 35.20524920370517, 32.21931456613152, 44.07717212042314, 48.56798614652851]

edited Nov 13 '19 at 21:35

answered Nov 12 '19 at 23:15

Konstantinos Katsantonis

1,280
11
17

Why call `str` before printing the result? – AMC Nov 12 '19 at 23:54
No list comprehension either? – AMC Nov 13 '19 at 21:37

score 0 · Answer 3 · answered Nov 12 '19 at 23:16

I think this should give you the solution you're after. The code below puts the low and high lists into a list of its own, and iterates through each index, randomly picking one.

from random import randint

low = [19.35, 18.09, 17.44, 16.46, 18.38, 20.78, 24.99, 25.22, 24.67]
high = [45.04, 44.05, 42.07, 41.95, 43.45, 46.04, 49.12, 50.94, 49.32]

lists = [low, high]
newlist = []
for i in range(0, len(low)):
    newlist.append(lists[randint(0, 1)][i])

This is an output I just generated, for reference:

[19.35, 18.09, 42.07, 16.46, 18.38, 46.04, 24.99, 25.22, 49.32]

I believe OP wants to generate a random number in the range `[a, b]`, not choose one of `a` or `b`. — AMC, Nov 12 '19 at 23:51

Python - Generating a New List Randomly From Two Other Lists

3 Answers3

A few comments on your code:

1)

2)

3)