freecodecamp project: building a markdown preview

started 11.43, finished 12.51

these were the requirements of the project

while doing the project, this link helped a lot (though i should really know by now how to do it - i’ll have to practice more jquery!):

• using jquery to update a div's content while typing automatically

end product

aaaaand this project is officially completed!

datacamp course: hypothesis testing in python

started at 16.49 and finished at 18.09

i realised i’ve been getting burnt out doing this track because of how much i despised statistics in a-level maths. now i know to make sure to pick a track i’m actually interested in and will enjoy completing, and not just because the certifications will be worth it (still gonna finish the track because i’m almost there already)

chapter 3: proportion tests

• one-sample proportion tests
  • p - population proportion (unknown population parameter)
  • p-hat - sample proportion (sample statistic)
  • p0 - hypothesised population proportion
• independence of variables
  • statistical independence - proportion of successes in response variable is same across all categories of explanatory variable
  • chi-square tests of variance are almost always right-tailed

chapter 4: non-parametric tests

• assumptions in hypothesis testing
  • every hypothesis test assumes each sample is a random subset sourced from its population
    • so, it is not representative of its population
  • every observation in the dataset is independent
    • so, increased chance of false negative/positive error
  • sample is big enough to mitigate uncertainty so central limit theorem applies
    • so, there may be wider confidence intervals and an increased chance of false negative/positive errors
• non-parametric tests
  • z-test, t-test and ANOVA are parametric tests. they assume a normal distribution and require sufficiently large sample sizes
  • non-parametric tests avoid parametric assumptions and conditions, using ranks of data
  • non-parametric tests are more reliable for small sample sizes and when data isn’t normally distributed
  • wilcoxon-signed rank test
    • works on ranked absolute differences between pairs of data
    • incorporates sum of ranks for negative and positive differences

back to busuu business?!!

i finally hopped back onto my language dailies, yes all three of them

french busuu - started at 19.01 and finished at 19.17. i studied 3.3 and 3.4.

german busuu - started at 19.22 and finished at 19.31. i studied 3.3 and 3.4.

japanese busuu - started at 19.31 and finished at 19.47. i studied 6.1, 6.2, 6.3 and 6.4.