I am not much of a math guy, but know some basics of pdfs, pmfs, Bayes theorems, probability distribution and stuffs. I am actually trying to build a Bayesian Network that models the personality of humans. For that I need independent probabilitiy distributions of different factors (such as age), and their effects, such as happiness.

The way in which I am collecting data is on the following Google Sheet. https://docs.google.com/spreadsheets/d/1OTMSUjalaT3K7sbwyfhvxooV65kI6X9umb6LCZDSUEk/edit#gid=0

My Question here is, Given the data above (in the gSheet) How Can I Calculate the pdf for P(Age), and conditional probability distribution of P(Age | Effects) where Effects = things like Happiness, Anger..

I'm like on a blackout on how to calculate these stuffs. Can anyone help? Please


You have a lot of learning to do. To begin, you can estimate $P(A|B)$ as the fraction of elements where B is true that also have A true. For example, $P(age=41|happiness=high)$ is approximately the ratio of the count of all members with age 41 and happiness level of high over the count of all those with happiness level high.

  • $\begingroup$ Is there, like a dynamic way to calculate for example: P(age = some_function_that_describes_how_age_varies | happiness = some_function_that describes_how_happiness_varies ) ? $\endgroup$ – cipher Jul 11 '15 at 17:52
  • $\begingroup$ It seems like your data is in spreadsheet format. I am not a spreadsheet expert, but you might want do check out the COUNTIF and COUNTIFS commands. If you move your data to a more powerful computing environment (MAPLE, Mathematica, R) you have more options $\endgroup$ – John McGee Jul 11 '15 at 17:57
  • $\begingroup$ The sheet is just to explain what I am trying to achieve. I could move my data to other envs. but Is the only way to do this is to make my data discrete? $\endgroup$ – cipher Jul 11 '15 at 18:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.