ScienceFair.jl

"""
@file      ScienceFair.jl

@author    Alex Matthew Jones-Martin
@copyright 2016 Gandalf Software, Inc.

@brief     My science fair project
"""
module ScienceFair
end

# This is a plotting package that will let me display
# my results in different ways
using Plots
pyplot()

# This lets me read a file with data separated by commas
using CSV
using DataStreams

# Open the file with the data I collected
myfile = CSV.Source(joinpath(homedir(), "sciencefair.csv"))

# Load the data from the open file
mydata = Data.stream!(myfile, Data.Table)

# Get the number of rows of data that I collected
num_rows = myfile.schema.rows

# Make vectors with all of the data from my file
vec_name   = Vector{String}(num_rows)
vec_digits = Vector{Int}(num_rows)
vec_age    = Vector{Int}(num_rows)

for i = 1:num_rows
    vec_name[i]   = mydata[i, 1].value
    vec_digits[i] = mydata[i, 2].value
    vec_age[i]    = mydata[i, 3].value
end

# Find all the different ages that data was collected for, sorted
# Julia makes this easy, with a function that returns a vector of only the unique entries
# in a vector, and then has a function that will return a vector in sorted order
ages = sort(unique(vec_age))

# Get the number of unique ages
num_ages = length(ages)

# Create a vector that stores all of the results for a particular age
digits_by_age = Vector{Vector{Int}}(num_ages)
for i = 1:num_ages
    # Start off with an empty vector
    digits_by_age[i] = Vector{Int}()
end

# For each unique age, store all results
for i = 1:num_ages
    age = ages[i]
    for j = 1:num_rows
        if age == vec_age[j]
            digits = vec_digits[j]
            # Add these results to the previous results for this age
            push!(digits_by_age[i], digits)
        end
    end
end

# Calculate different results, such as mean (average), median, and range, for each unique age
mean_by_age   = Vector{Float64}(num_ages)
median_by_age = Vector{Float64}(num_ages)
low_range     = Vector{Int}(num_ages)
high_range    = Vector{Int}(num_ages)
for i = 1:num_ages
    digits = digits_by_age[i]
    mean_by_age[i]   = mean(digits)
    median_by_age[i] = median(digits)
    low_range[i]     = minimum(digits)
    high_range[i]    = maximum(digits)
end

# Plot data (with help by Tom Breloff, a friend of my dad)
plot(ages, mean_by_age, lab="Mean", m=(10, :diamond), err=(mean_by_age - low_range, high_range-mean_by_age), xlab = "Age", ylab="Correct Digits") 
plot(ages, median_by_age, lab="Median", m=(10, :square))
#end