Getting the same results

[YA_10963798]

Although I'm using different variables, I don't know why I'm having the same results for Mi .. 

It is on page 14 

Thanks in advance 

[Werner_E · Accepted solution]

You forgot to attach the excel data file. The one you posted earlier does not have a sheet #28.

Maybe you are looking for

as I don't think that you expect a nested vector of vectors, which you would get if you omit the index i on dz. Vectorization isn't necessary anyway.

EDIT: An easier way to achieve the same result would be

but it fails because vector dz has one element less than the other two vectors involved.

So you would have to make all three involved vectors equal in length, either by getting rid of the last elements in f.1... and q.cor... using submatrix (this gives the very same result as the program shown) or add an additional 0,01m at the end of dz (which results in in a vector M2 one element larger).

[YA_10963798]

Thank you Werner

[StuartBruff]

No problem, Yusra. Werner had given you the answer and I’d meandered the rest of the thread off into something totally unrelated, anyway. 🙂

Stuart

[StuartBruff]

Hello, Yusra.

Could you please upload your latest version of "CPT input gef.xlsx"?  The version I've got creates errors when it tries to read range "[28]A2:D2196".

However, looking at your program, I'm wondering why you return M_i instead of M? 

Stuart

[Werner_E]

For testing purposes you may replace [28] by [27] in range6 or delete range6 altogether

[StuartBruff]

Deleting Range 6 (or, rather, ignoring it by changing the i_n range) didn't work for me.  There are references to it later in the worksheet, and I felt that it wouldn't be a good use of my time to hunt for any other potential errors that might pop up.

I'll try using 27.  Sounds good.

Stuart

[StuartBruff]

Yes, using 27 worked.  However, it would still be nice to have the version of the Excel workbook to see what else might pop up.

I came to the same conclusion as you, which I already suspected I would based on the screenshot.  Worksheet success seems to correlate with the weather - it's stopped raining. I shall have to keep a record of the weather at the time my worksheets work! 😁

Stuart

[Werner_E]

---

@StuartBruff wrote:

Yes, using 27 worked.  However, it would still be nice to have the version of the Excel workbook to see what else might pop up.

I came to the same conclusion as you, which I already suspected I would based on the screenshot.  Worksheet success seems to correlate with the weather - it's stopped raining. I shall have to keep a record of the weather at the time my worksheets work! 😁

Stuart

---

Oh yes, it is certainly worth initiating an intensive and thorough statistical study in this regard 🙂

[StuartBruff]

---

@Werner_E wrote:

---

@StuartBruff wrote:

Yes, using 27 worked.  However, it would still be nice to have the version of the Excel workbook to see what else might pop up.

I came to the same conclusion as you, which I already suspected I would based on the screenshot.  Worksheet success seems to correlate with the weather - it's stopped raining. I shall have to keep a record of the weather at the time my worksheets work! 😁

Stuart

---

Oh yes, it is certainly worth initiating an intensive and thorough statistical study in this regard 🙂

---

It's funny you should say that ...

One of the many projects on my "tomorrow" list is checking accuracy and variability of various temperature sensors.  I have several different sensors, with various claims of accuracy, that I intend to hook up to a Raspberry Pico to get some actual data.  In addition to temperature, I also intend to collect relative humidity and air pressure data.  My biggest problem is finding a decent (affordable) calibrated sensor to give an accurate reference to check the other sensors against.  And then there's ensuring that they are all experiencing the same environmental conditions.

It would be nice to have a Mathcad worksheet that got the testbed results live and updated the stats.  Generating tabular data isn't a problem and I can press F5 to manually force a Mathcad read of the file. Unless I start poking around with C/C++/C# programming ... or maybe an Advanced Control ... I don't know how to automate the reading.

I suppose I now need an IOT button next to my laptop so I can notify the system when I write something in Mathcad that actually works.

According to the various sensors I currently have scattered about the place, my home temperature range is (14.6, 13.1, 11.9, 12.6, 14.9, 12.0) °C.  The last temp (12.0 °C) is according to Alexa; the rest are from a WiFi sensor and a weather station with an internal sensor and three external sensor units.  The Alexa sensor (in an Echo Dot speaker) is next to one of the weather station sensors (which read 11.9 °C). 

The other sensors are all within 50 cm of each other yet have noticeable differences - the 12.6 °C and 14.9 °C sensors are next to each other.  

I did keep an Excel workbook holding readings from another weather station, Weather Underground, and Glasgow Airport.  From about 90 readings, I found the following.

Locn	WU Temp Diff	WU RH Diff	WU Pressure Diff	GA Temp Diff	GA RH Diff	GA Pressure Diff
Max	2.9	6.0	-4.6	5.9	10.0	-4.6
Min	-2.5	-18.0	-7.0	-1.9	-19.0	-7.0
Stdev	1.4	5.8	0.5	1.7	6.9	0.6
Mean	0.7	-6.8	-5.6	1.4	-7.9	-5.7

[Werner_E]

Looks like there are quite some differenced between the two external weather stations WU and GA, too, when it comes to temperature and humidity?

[StuartBruff]

---

@Werner_E wrote:

Looks like there are quite some differenced between the two external weather stations WU and GA, too, when it comes to temperature and humidity?

---

Indeed.  I became interested in the subject after spotting discrepancies between multiple weather stations that should be reporting the same measurement results.  However, there are many reasons why they might differ, even for stations close to each other.   

. 

One reason is that location and housing can have effects on what the sensed air temperature is.  I know, having worked on one, that UK military and UK Met Office airfield weather stations are calibrated and housed to minimise such things as solar heating of the sensors.  Microclimate can have significant effects even when equipment is closely located; local topography (hills, rivers, lakes, buildings) can make noticeable differences (cities are usually a couple of degrees warmer than the surrounding countryside). I've just looked at a couple of weather stations on the island of Anglesey - Royal Air Force Valley and a nearby site, and there are reading differences; Of note, RAF Valley is next to the sea whilst the other site is a couple of miles further inland.  The Met Office states:

Observations made at synoptic stations should represent the wider area around the station and not be unduly influenced by local effects. The ideal site should be on level ground with no trees, buildings or steep ground nearby that might influence the measurements. Undesirable influences include the warming effect of buildings on the measurement of temperature and the sheltering or shading effects of trees on the measurement of sunshine and wind. Wherever possible, a station should not be located in a frost hollow where overnight temperatures on still clear nights may be far lower than at neighbouring locations. In a similar way, winds measured at the top of a hill or steep escarpment will be unrepresentative of the wider area.

Taken from https://www.metoffice.gov.uk/weather/guides/observations/uk-observations-network.   The airfield weather stations follow these guidelines as closely as possible and the weather stations are usually enclosed in a vented box to protect the equipment from direct solar radiation - it's only the air properties that really matter.  More information about weather stations can be found at https://www.metoffice.gov.uk/weather/learn-about/how-forecasts-are-made/observations/weather-stations. 

Weather Underground has guidance on buying and installing a weather station and connecting it to their network: https://www.wunderground.com/pws/about.  Weather Underground bases it weather observations for my locale from an airport 22 miles (35 km) away that is in flat terrain and inland from the sea - I'm about 400 metres from the sea and have steep hills about 200 meters behind me.  

It's off this thread's topic ... thread drift seems to be quite common ... but meteorology is potentially of interest to Mathcadders who look to improve their data analysis capability.

Stuart

[Werner_E]

It is indeed interesting, but also sobering, how seemingly simple things such as temperature measurement become quite complex and time-consuming when you take a closer look.

[StuartBruff]

---

@Werner_E wrote:

It is indeed interesting, but also sobering, how seemingly simple things such as temperature measurement become quite complex and time-consuming when you take a closer look.

---

Indeed.  One of these days I'm going to start listening to myself when the Experience neural net says, "It's a rabbit hole. You do not want to go down there even if rabbits are tasty".  But i say that every time, listing all my priorities, but jump down it like a ferret on a sugar high, anyway.   I think it goes back to my childhood experience of wondering how I was supposed to measure things exactly when the markings on the rulers were so wide and blurry at the edges, and then checking all the rulers I could find to see which, if any, gave an answer.

Ah, the good old wooden ruler - so useful for teacher's about to deliver administrative punishment to a recalcitrant school student.  It's funny how the boys always got a good whacking (hand for minor mischief, backside for above and beyond trouble), but the girls were just lightly admonished.

Stuart

[LucMeekes]

"It would be nice to have a Mathcad worksheet that got the testbed results live and updated the stats."

That's where good old MathConnex should be able to help....

Success!

Luc

[StuartBruff]

I'd forgotten all about Mathconnex, Luc!

Stuart

[YA_10963798]

Hi Stuart, 

Sorry I just saw this message . Thank you so much for trying to help.