** UnhandledExceptionEventHandler :: OFFICIAL LC / CHATTER THREAD ** - Page 688 - Computer Technical Help

Two Plus Two Forums Other Topics Computer and Technical Help

** UnhandledExceptionEventHandler :: OFFICIAL LC / CHATTER THREAD **

Post Reply Subscribe

...

Page 688 of 1603

First

188 488 588 638 668 678 683 684 685 686 687 688 689 690 691 692 693 698 708 738 788 888 1188

Last

Page 688 of 1603

First

188 488 588 638 668 678 683 684 685 686 687 688 689 690 691 692 693 698 708 738 788 888 1188

Last

03-16-2015 , 09:02 PM

#17176

gaming_mouse

Carpal \'Tunnel

Join Date: Oct 2004 Posts: 13,786

Quote:

Originally Posted by e i pi

Anyone here have any experience with SMS api's/services? I recognize the name twilio and they look legit but is there any other company worth looking into? I don't trust google searches for product reviews anymore so thought I'd ask here.

twilio is great. i think they're the biggest player in the market, will be around for a long time, good docs, transparent pricing.

Quote

$25m Guaranteed WPM on CoinPoker

Join the action now

Daily Rewards • Splash Pots • CoinRaces

03-17-2015 , 04:08 AM

#17177

jmakin

banned

Join Date: Jan 2008 Posts: 27,798

i am so burnt out on school. 3 more years probably. and i feel like I've already done so much.

better be worth it

Quote

03-17-2015 , 08:29 AM

#17178

ChrisV

Carpal \'Tunnel

Join Date: Jul 2004 Posts: 40,336

Quote:

Originally Posted by e i pi

This is exactly the business I'm in, but Australia only, so useless presumably.

I would advise against going for anything ludicrously cheap, they will be using so called "gray routes" that skirt the borders of legality. Either your customers will find their number/messages scraped for spam, or you will get dubious results. In the worst cases, one supplier reported to my company that messages had been delivered when they actually hadn't. This is obviously a huge problem as there's no way of assessing the extent of it or when it has been fixed.

Quote

03-17-2015 , 10:12 AM

#17179

Anais

Carpal \'Tunnel

Join Date: Mar 2006 Posts: 6,663

Is DeVry one of those scammy Bull**** schools or is it legitimate?

I could have sworn at one point it was just a useless diploma mill, but maybe I'm thinking of another school.

Quote

03-17-2015 , 10:44 AM

#17180

candybar

Pooh-Bah

Join Date: Aug 2011 Posts: 4,375

Quote:

Originally Posted by Larry Legend

candybar,

Financial software? Traders?

Yes & no, though I've done that as well. Long before I got serious about software engineering, I once leveraged my programming skills to get a junior position on a trading desk. Which lasted just a few months but it made me question my faith in humanity. Sell-side traders are not very nice people to work with. Bankers and buy-siders at least wear sheep's clothing.

Quote

03-17-2015 , 03:18 PM

#17181

Barrin6

Carpal \'Tunnel

Join Date: Dec 2005 Posts: 6,948

Quote:

Originally Posted by Anais

Is DeVry one of those scammy Bull**** schools or is it legitimate?

I could have sworn at one point it was just a useless diploma mill, but maybe I'm thinking of another school.

I think it's more of a tech trade school. We had a sysadmin who graduated from there.

Quote

03-17-2015 , 10:54 PM

#17182

daveT

S.A.G.E. Master

Join Date: Jun 2005 Posts: 23,955

Okay, I'm stumped.

I found this problem:
https://www.interviewcake.com/questi...te-5-sided-die

And I was kind of surprised at the answer. So, I created this entirely naive "program" that I would think computes the correct result, but it doesn't compute the correct answer.

I originally had looping, but since it gave the wrong results, I wrote it out like this:

Code:

import random

def r7to5(r):
## 7 / 5 = 1.4
    if 0 <= r < 1.4:
        return 1
    if 1.4 <= r < 2.8:
        return 2
    if 2.8 <= r < 4.2:
        return 3
    if 4.2 <= r < 5.6:
        return 4
    if 5.6 <= r <= 7.0:
        return 5

def rand7():
    return random.randint(0, 7)

def rand5():
    r = rand7()
    return r7to5(r)

d = {}
for i in range(1001):
    a = rand5()
    if a in d:
        d[a] += 1
    else:
        d[a] = 1

print(d)

The output:

Code:

>>> {1: 243, 2: 130, 3: 245, 4: 124, 5: 259}
>>> {1: 252, 2: 118, 3: 244, 4: 123, 5: 264}
>>> {1: 242, 2: 128, 3: 252, 4: 130, 5: 249}
>>> {1: 263, 2: 121, 3: 243, 4: 120, 5: 254}
>>> {1: 265, 2: 146, 3: 235, 4: 121, 5: 234}
>>> {1: 248, 2: 114, 3: 255, 4: 118, 5: 266}
>>> {1: 250, 2: 138, 3: 255, 4: 126, 5: 232}
>>> {1: 236, 2: 123, 3: 245, 4: 121, 5: 276}
>>> {1: 248, 2: 123, 3: 212, 4: 142, 5: 276}
>>> {1: 264, 2: 128, 3: 241, 4: 129, 5: 239}
>>> {1: 254, 2: 137, 3: 244, 4: 117, 5: 249}

What am I missing here?

Quote

03-17-2015 , 11:01 PM

#17183

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

You're looking for 0 to 7 instead of 1 to 7?

Quote

03-17-2015 , 11:10 PM

#17184

blackize5

old hand

Join Date: Sep 2012 Posts: 1,358

2 and 4 from r5to7 are the only results that don't correspond to a range including two integers. Rand 7 is giving you an integer-0 to 7 and you're looking for values to fall between, for example, 2.8 through 4.2 which covers both 3 and 4. So 3 or 4 from rand would both map to 3

Hopefully this makes sense I'm on my phone now

Quote

03-17-2015 , 11:21 PM

#17185

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

I like blackize5's answer better. Couldn't you just loop until you get a number < 6 from the rand 7 function?

Quote

03-17-2015 , 11:24 PM

#17186

fredd-bird

Carpal \'Tunnel

Join Date: Aug 2007 Posts: 19,359

The question doesn't mention a random integer, just a random number. So I changed it to random.uniform to output a float. Then I changed the ranges in your if statement to match the updated range of possible outputs from rand7().

Code:

def r7to5(r):
## 7 / 5 = 1.4
    if 1.0 <= r < 2.2:
        return 1
    if 2.2 <= r < 3.4:
        return 2
    if 3.4 <= r < 4.6:
        return 3
    if 4.6 <= r < 5.8:
        return 4
    if 5.8 <= r <= 7.0:
        return 5

def rand7():
    return random.uniform(1, 7)

Code:

{1: 212, 2: 217, 3: 196, 4: 169, 5: 207}
{1: 214, 2: 182, 3: 205, 4: 192, 5: 208}
{1: 223, 2: 192, 3: 176, 4: 202, 5: 208}
{1: 219, 2: 210, 3: 205, 4: 187, 5: 180}

Quote

03-17-2015 , 11:29 PM

#17187

daveT

S.A.G.E. Master

Join Date: Jun 2005 Posts: 23,955

ah, sheesh. I forgot to change randint.

The assertion from the site is this:

This kind of math is generally outside the scope of a coding interview, but: if you know a bit of number theory you can prove that there exists no solution which is guaranteed to terminate. Hint: it follows from the fundamental theory of arithmetic..

Is it saying that there is no guarantee of termination using recurrences or am I missing something else?

Quote

03-17-2015 , 11:34 PM

#17188

blackize5

old hand

Join Date: Sep 2012 Posts: 1,358

The solution uses a loop to throw out any value greater than 5 trying again until it gets 5 or less. In theory the rng could never return a number 5 or less.

Nvm the floating point stuff. It's implies that rand7 returns integers 1 to 7

Quote

03-17-2015 , 11:37 PM

#17189

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

Doesn't seem like much of a RNG if it's always 6 or 7... The hint about not caring about run time backs that up though.

Quote

03-18-2015 , 12:02 AM

#17190

e i pi

veteran

Join Date: Feb 2010 Posts: 2,230

it's a more interesting question with random integers imo.

also apparently you can use the modulus operator with floats in python.

>>> 5.25 % .5
0.25
>>> 5.25 % .20
0.04999999999999971

Quote

03-18-2015 , 12:09 AM

#17191

e i pi

veteran

Join Date: Feb 2010 Posts: 2,230

>>> def rand5():
... return sum([randint(1,7) for x in range(1,5)]) % 5 + 1

I think this works for integers

Quote

03-18-2015 , 12:11 AM

#17192

Craggoo

banned

Join Date: Sep 2006 Posts: 12,645

I'm about to start applying for web dev jobs. I have a degree in criminal justice. Should I include that in my resume? I'm trying to avoid including irrelevant information. Am I wrong in thinking that a college degree in an unrelated field is irrelevant information?

Quote

03-18-2015 , 12:13 AM

#17193

Dudd

Carpal \'Tunnel

Join Date: Sep 2004 Posts: 54,153

I don't think you can use the floating point versions because the final range is closed on both ends, while the others have one open end and one closed end. You'd need to have infinite precision for those ranges to have equal length, but then there would be an infinite number of random numbers to choose from, so how could a random number generator choose one? You could leave the final range open too, but then whenever rand7() returned exactly 7.0, you'd have to reenter rand7() to get a new value just like in the integer version, so you'd have a worst case infinite scenario again, albeit a much, much less likely infinite scenario

Quote

03-18-2015 , 12:17 AM

#17194

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

Quote:

Originally Posted by e i pi

>>> def rand5():
... return sum([randint(1,7) for x in range(1,5)]) % 5 + 1

I think this works for integers

I was thinking that too but turns out there is mod bias, mod 5 returns an even remainder for 1 - 7.

Quote

03-18-2015 , 12:22 AM

#17195

adios

Carpal \'Tunnel

Join Date: Sep 2002 Posts: 22,558

Quote:

Originally Posted by Craggoo

Take this FWIW as it is strictly an opinion. I would just mention it where you show your education unless you really don't want to have any discussion about it in an interview. I say include it because it shows your determination in acquiring development skills as you were motivated to learn. I could definitely be way off base here though. It may be worth the money to consult with a pro who does resumes for web dev jobs.

Quote

03-18-2015 , 12:26 AM

#17196

blackize5

old hand

Join Date: Sep 2012 Posts: 1,358

I'd say have two versions. Prefer the one without unless the job posting requires a degree.

The way I see it, it's never going to be a positive beyond getting past the resume filter. And I suspect there's a little stigma for criminal justice degrees.

Quote

03-18-2015 , 12:28 AM

#17197

e i pi

veteran

Join Date: Feb 2010 Posts: 2,230

Quote:

Originally Posted by kerowo

I was thinking that too but turns out there is mod bias, mod 5 returns an even remainder for 1 - 7.

yea but I added the values of 5 different rand7 calls. So it's like taking mod 5 of 5-35 so I think the distribution isn't biased.

Last edited by e i pi; 03-18-2015 at 12:42 AM.

Quote

03-18-2015 , 12:30 AM

#17198

daveT

S.A.G.E. Master

Join Date: Jun 2005 Posts: 23,955

I don't know number theory very well, but I suspect that, given a truly random sequence, at least one would have to have an infinite string of 7s.

After thinking on it a bit more, I don't think there is a closed solution to this problem if we are dealing with prime integers. Eh, I misinterpreted the initial problem.

Quote

03-18-2015 , 12:33 AM

#17199

daveT

S.A.G.E. Master

Join Date: Jun 2005 Posts: 23,955

Code:

def rand5():
    return sum([random.randint(1,7) for x in range(1,5)]) % 5 + 1


d = {}
for i in range(10001):
    a = rand5()
    if a in d:
        d[a] += 1
    else:
        d[a] = 1

print(d)

Code:

>>> {1: 1999, 2: 2006, 3: 1969, 4: 2020, 5: 2007}
>>> {1: 1990, 2: 2046, 3: 1986, 4: 1996, 5: 1983}
>>> {1: 1985, 2: 2054, 3: 1996, 4: 2028, 5: 1938}
>>> {1: 2004, 2: 1988, 3: 2067, 4: 1987, 5: 1955}
>>> {1: 1916, 2: 2050, 3: 2052, 4: 1978, 5: 2005}
>>> {1: 2063, 2: 2020, 3: 1957, 4: 1950, 5: 2011}
>>> {1: 2030, 2: 1978, 3: 2027, 4: 1998, 5: 1968}
>>> {1: 2022, 2: 2041, 3: 1987, 4: 2010, 5: 1941}
>>> {1: 1995, 2: 2028, 3: 1973, 4: 2027, 5: 1978}
>>> {1: 1963, 2: 1983, 3: 1979, 4: 2032, 5: 2044}
>>> {1: 1989, 2: 1993, 3: 2009, 4: 1969, 5: 2041}

or 1000000 trials:

Code:

>>> {1: 200388, 2: 201539, 3: 200478, 4: 198936, 5: 198660}
>>> {1: 200440, 2: 201216, 3: 199791, 4: 199375, 5: 199179}

Quote

03-18-2015 , 12:51 AM

#17200

e i pi

veteran

Join Date: Feb 2010 Posts: 2,230

It takes more programming skills to verify solutions than to find them for these types of problems.

2000th post

Quote

$25m Guaranteed WPM on CoinPoker

Join the action now

Daily Rewards • Splash Pots • CoinRaces

Page 688 of 1603

First

188 488 588 638 668 678 683 684 685 686 687 688 689 690 691 692 693 698 708 738 788 888 1188

Last

Post Reply Subscribe

...

Page 688 of 1603

First

188 488 588 638 668 678 683 684 685 686 687 688 689 690 691 692 693 698 708 738 788 888 1188

Last