MetabolicStat
This software is not affiliated with, sponsored by, or compensated by any health care industry company or glucose monitoring company. Neither Keto-Mojo or Nutrisense companies have contributed to this work or provided any renumeration. This software has nothing to do with them.
Furthermore:
a> There is nothing expressed or implied about this software's fitness of purpose, even the purpose of analyzing the intended statistics for which it is written.
b> This software offers no medical advice. This software offers no diet advice. This software does not provide warnings or any judgments of any data.
c> There are no guarantees expressed or implied. This software is made available AS IS.
d> I have no conflicts of interest by or for any company where that company is actively perusing wellness programs.
e> This software is based on a very old linear regression algorithm I originally used on a TI-59 calculator in 1978. This statistical algorithm was apparently published in the Galton 1890's.
f> None of the ‘medical’ data in this project represents private patient information protected by HIPAA. It is all test data.
Now that you have collected your Glucose and ketone data for a few months, (Congratulations you now know that the answer to your question is 42 :)) how can you make sense of it and track your progress? (i.e., What exactly was the question you wanted to answer anyway?)
This project is designed to help you find that answer, creating meaningful linear regressions of 'value vs 'time', and allows the user to create detailed graphs with excel or other graphing tools, where you can explore your data with different types of graphs, run statistical analysis, and print reports to discuss with friends.
Besides an external spread sheet or other data graphing program there are no other tools required. Linear regression statistical math was published in journals in the 1890s’ the software class to calculate this statistic is included in this application with documentation in that class.
This project loads csv glucose and ketone data and computes linear regressions vs time of GKI, Ketone, and Glucose for analysis and graphing purposes.
The purpose of doing this analysis is to enable the user to examine their glucose and ketone data and search for relationships and patterns.
This process requires user involvement. This software offers no advice or medical opinions on the meaning of different values. There are no warning messages displayed or printed.
Your glucose/meter sensors may give you readings that look concerning to you. In that event you should contact your doctor or go to urgent care. This software just crunches numbers. It makes no judgements. It gives no advice. It contains no recipes and does not recommend a diet.
Glucose (ketone) data set can quickly grow very large. This is particularly true when a CGM is used to collect the data. Various companies who provide a meter or CGM typically provide a data service to help customers ‘grock’ their data. If you wanted to graph years’ worth of data there could easily be over 15 thousand records. This program selects data from an input data archive file, sorts into histogram bins as a series of linear-regressions, and prints out the results for each linear regression.
Example:
| date | name | meanx | minx | maxx | sdx | slope | N |
|---|---|---|---|---|---|---|---|
| 8/28/2020 | CGM-7/1/2020-8/31/2020 | 85.69 | 52 | 111 | 0.5928 | .0175 | 255 |
| 8/31/2020 | MGL-8/31/2020-10/31/2020 | 80.69 | 40 | 114 | 0.1618 | -.5091 | 5271 |
| 10/31/2020 | MGL-10/31/2020-12/31/2020 | 89.91 | 51 | 161 | 0.1615 | .16984 | 4936 |
| 12/29/2020 | MGL-12/31/2020-3/2/2021 | 84.46 | 36 | 140 | 0.1309 | -.20717 | 5757 |
| 3/1/2021 | MGL-3/2/2021-5/1/2021 | 86.86 | 44 | 276 | 0.1588 | .2588 | 5163 |
| 10/17/2020 | MGL-5/1/2021-7/1/2021 | 82.73 | 35 | 159 | 0.185 | -.01413 | 5876 |
| 12/8/2020 | MGL-7/1/2021-8/31/2021 | 85.94 | 51 | 196 | 0.1784 | .01284 | 5106 |
| 8/31/2021 | MGL-8/31/2021-10/31/2021 | 85.24 | 51 | 178 | 0.1582 | .31448 | 4829 |
| 9/4/2021 | MGL-10/31/2021-12/31/2021 | 77.51 | 36 | 102 | 0.1639 | .1219 | 5309 |
| 12/31/2021 | MGL-12/31/2021-3/2/2022 | 80.73 | 35 | 104 | 0.186 | -.04755 | 4878 |
| 3/2/2022 | MGL-3/2/2022-5/2/2022 | 79.11 | 35 | 169 | 0.1625 | -.0 | 4916 |
(under construction)
This current version writes the statistics *.csv files to the same folder/directory where the data file resides.
I am using it this way:
Command line, bashsh, sh, or powershell
Create a folder easy to access such as c://MetabolicData/ create a repo clone of this source and do a ''' git pull repo ''' Use a version of visual studio to build the exe file.
Go to your targe folder where you build the software and run the executable against the target data base file.
The *.csv files will be created alongside that data file.
The report printed on the console is a trace of the programs effort to generate the different *.csv tables.
After the *.csv files are created they can be imported into excel or other post process to visualize the data.