JonasMoss

First Reading

On the Consistency Rule in Causal Inference Axiom, Definition, Assumption, or Theorem? (Pearl, 2010, 4 page) One of the big problems with the causality literature is the terminology and the lack of foundationas for everyone to agree on.(Think about a vector space -- everyone agrees what it is. That's where we want to be.) The consistency rule appears to me to be the corner-stone of an axiomatic development of causality theory.

The following papers are mentioned in the Pearl paper and are a part of the assignment:

1. The consistency statement in causal inference: a definition or an assumption

2. Concerning the consistency assumption in causal inference

JonasMoss

JonasMoss

JonasMoss

from collections import Counter
def f(i, s):
    total = np.cumsum([0] + list(Counter(coding(s)).values()))
    total = total / total[-1]
    values = np.diff(total)
    return values[i-1]

coding(data_student['apply'],f

JonasMoss

### Here's some more info about default values! 
### Especially when they are difficult to understand.

### An easy example of default values.
def f(x, a = True): 
    if a:
        return (x + 1)
    if not a:
        return (x + 2)

JonasMoss

#' Sample from a kernel density with the Epanechnikov kernel.
#' 
#' @param n Number of observations to sample.
#' @param x The data from which the estimate is to be computed.
#' @param bw Desired bandwidth.
#' @return A numeric vector with n sampled data points from the kernel
#' density estimator.
#' @details The factor sqrt(5) in is needed since the standard deviation of 
#' the kernel is 1/sqrt(5). This makes the normal kernel and the Epanchnikov
#' kernel comparable.