Comments for Juan Carlos Méndez-García

Comment on El Juego de la Cerveza - Systems Dynamics by Vibeke

Vibeke — Thu, 14 Aug 2008 16:49:06 +0000

Hello,
Great material, but do you have it in English ?

Kind regards,
Vibeke (Norway)

Comment on Wireless access on my Macbook Pro by Juan Carlos Méndez-García » Blog Archive » Connecting Mac OS X to the net through a Treo 680 GPRS

Fri, 01 Aug 2008 10:26:01 +0000

[...] Click on Advanced. In my case, the Treo worked fine with the Generic script as shown below. If not, find out what works for your phone. For instance, for a RAZR V3xx I used in the US, Ross Barkman’s scripts worked well, as mentioned in a post some time ago. [...]

Comment on Modeling market adoption in Excel with a simplified s-curve by Debbie

Debbie — Thu, 24 Jul 2008 13:57:13 +0000

Hi Juan,
Thanks for providing this example. It’s very useful! What I would like to do, however, is to find the gradients of the S-Curve at any desired point. How would I find that using excel, after my Curve has been created?
Thanks!

Comment on Math on the simplified market adoption s-curve for Excel by Juan Carlos Méndez-García

Juan Carlos Méndez-García — Thu, 24 Jul 2008 06:40:59 +0000

@Edward - I have trouble understanding your problem description.
“during the life of the project, it is estimated to use 45GB of space” — Are you going to add data each month data[m] so SUM(data[m] for m=1..24) ~= 45GB, right?
Then, Growth is NOT =f(m) as you tried.

If I understand correctly, total_data[m] = f(m), where f is my formula with parameters saturation=45 [GB], hypergrowth=4, takeover=14.

It is somewhat counter intuitive to define takeover in terms of a takeover_rate applied over a growth table. takeover is simply an assessment of when you believe the technology is “mature” or “quite adopted”.

With that you would have an S-shaped curve that grows from 0GB to ~45GB by month 24.

The rest of your variables, you would derive backwards, and it would be something like
data[m]=growth[m]=total_data[m]-total_data[m-1] assuming total_data[0]=0
growth_pct[m]=growth[m]/total_data[m] if I define growth_pct as the percentage by which the data grows each month versus the accumulated data. Other ways to define growth (month over month, etc) can be also derived backwards very easily.

However, that doesn’t jive with your other statements “What I need, is the month Growth total never to exceed estimated total size (45GB) and total Growth % to equal 100%. Make sense?”

Nope. I’m confused.

Is 45GB then the maximum allowable growth for a month? That would be described as follows:
For any i in 1..24 data[i]< =45GB, which is quite different from SUM(data[m] for m=1..24) ~= 45GB

With that formulation, it explains that you got ~151GB for total space.

Two more questions...
"total Growth % to equal 100%"... at which month?
Why you chose saturation = 25%

Hope it helps. I would recommend to check http://en.wikipedia.org/wiki/System_dynamics to help you clarify which variables you want to model as “stocks” and which ones are “flows”

Comment on Math on the simplified market adoption s-curve for Excel by Edward

Edward — Mon, 21 Jul 2008 18:57:55 +0000

Thanks a lot… I worked on the formula over the weekend and figured it out.

However, I’m now having one other problem… hope you don’t mind looking at.
————————
Let’s say you have a project starting on 1/1/07 and during the life of the project, it is estimated to use 45GB of space. I want to calculate monthly growth and track total growth. My table has 3 columns (Month, Growth, % Growth). Month should be 1 - 23, Growth = % * Data_Volume, and % Growth = f(month).

I’ve tried several different things:
1. Calculating % Growth = f(month) - ends up getting 151GB & 335% for total space.
2. Calcuating % Growth = current f(month) - prior f(month) - now get 11GB and 24% for total space.

What I need, is the month Growth total never to exceed estimated total size (45GB) and total Growth % to equal 100%. Make sense?

Thanks for all your help.

Conditions:
1. Base_month = 1
2. # of Project Months = 24
3. Saturation = 25%
4. Hypergrowth = 4 (=ROUND(takeover*0.3,0))
5. Taekover = 14 (=ROUND((MAX(Monthly_Growth_Table[Month])*takeover_rate),0))
6. takeover_rate = 60%
7. Data_Volume = 45GB

Comment on Harmonic Averages by Juan Carlos Méndez-García

Juan Carlos Méndez-García — Fri, 18 Jul 2008 10:15:20 +0000

More harmonic average examples on http://www.nerdblog.com/2008/01/harmonic-mean-mpg-standards.html

Comment on Gearing up for summer classes in Colombia by Juan Carlos Méndez-García » Blog Archive » El Juego de la Cerveza - Systems Dynamics

Fri, 18 Jul 2008 10:02:31 +0000

[...] of the activities conducted during the System Dynamics class in Colombia was the Beer Game - El Juego de la [...]

Comment on Harmonic Averages by Juan Carlos Méndez-García

Juan Carlos Méndez-García — Fri, 18 Jul 2008 05:48:49 +0000

@Mathias - You are absolutely correct. I’ll fix the formula on the post to avoid confusion. Thanks very much for the Excel function. As you point out, Excel’s built-in HARMEAN() does not allow for weights.

Comment on Math on the simplified market adoption s-curve for Excel by Juan Carlos Méndez-García

Juan Carlos Méndez-García — Fri, 18 Jul 2008 05:41:02 +0000

@Edward The model works regardless the time units, as long as everything is consistent. Let's say saturation is 25, hypergrowth is month 8, and takeover is month 28. So when you use the same formula f(month)=saturation/(1 + 81^((hypergrowth + takeover/2 - month)/takeover)) you would have for different months


Month   f(Month)
5 1.62
10 3.30
15 6.25
20 10.55
25 15.39
30 19.46
35 22.12

Comment on Math on the simplified market adoption s-curve for Excel by Edward

Edward — Thu, 17 Jul 2008 19:40:26 +0000

Juan,

Your example shows in years. However, I need to do my calculations in months, tried to convert, but doesn’t work correct. Any suggestions?

Month	f(Month)
5	1.62
10	3.30
15	6.25
20	10.55
25	15.39
30	19.46
35	22.12