tpapp

using AMDB
using DataStructures
using GZip
using UnicodePlots
using IntervalSets

# if we want to pass around code, extend it with your own machine name
# (gethostname# ()) and path
AMDB_path = get(Dict("tamas" =>
                     "/home/tamas/research/AMDB/data/AMDB_subsample.jls.gz"),

tpapp

using AMDB
using DataStructures
using UnicodePlots
using IntervalSets              # need to Pkg.checkout the latest version

# i made a new function to read the data, replace path for your own machine
records = deserialize_gz(expanduser("~/research/AMDB/data/AMDB_subsample.jls.gz"))

# suppose we are counting in 2005
date_interval = Date(2005,1,1)..Date(2005,12,31)

tpapp

using AMDB
using DataStructures
using IntervalSets

"""
Call `f` with each AMP spell in `records`. Useful for accumulators.
"""
function traverse_AMP_spells(f, records)
    for data in values(records)
        foreach(f, data.AMP_spells)

tpapp

using AMDB
using DataStructures
using IntervalSets
using Plots

plotlyjs()               # I like this backend, feel free to change it

records = deserialize_gz(expanduser("~/Documents/Thesis/AMDB_subsample.jls.gz"))

function spell_durations_in_year(records, interval)

tpapp

using AMDB
using DataStructures
using IntervalSets
using Plots
using DataFrames                # you may need to add these libraries
using StatPlots                 # with Pkg.add

records = deserialize_gz(expanduser("~/research/AMDB/data/AMDB_subsample.jls.gz"))
plotly() # this makes plots appear in your browser, you can use other backends

tpapp

\documentclass{beamer}

\usepackage{amsmath}
\usepackage{amsfonts}
\usepackage{booktabs}
\usepackage{mathtools}
\mathtoolsset{showonlyrefs}

%% no idiotic nav buttons
\setbeamertemplate{navigation symbols}{}

tpapp

######################################################################
# MWE for tabulating large CSV files, using plain Julia and IterateTables
# See the discussion at
# https://discourse.julialang.org/t/iterate-delimited-file-as-namedtuples/3616/3
#
# Some simplifications:
# - using the index of iterated tuples to make it comparable,
#   but would prefer to use the column name in real code.
# - not using compressed files (because CSV does not support Zlib streams ATM)
######################################################################

tpapp

"""
Lag-`k` autocorrelation of `x` from a variogram. `v` is the variance
of `x`, used when supplied.
"""
function ρ(x, k, v=var(x))
    x1 = @view(x[1:(end-k)])
    x2 = @view(x[(1+k):end])
    V = sum((x1 .- x2).^2)/length(x1)
    1-V/(2*v)
end

tpapp

# consistent random numbers
srand(UInt32[0xfd909253, 0x7859c364, 0x7cd42419, 0x4c06a3b6])

"""
    err(x, [prec])

Return two values, which are the log2 relative errors for calculating
`log1p(x)`, using `Base.log1p` and `Base.Math.JuliaLibm.log1p`.

The errors are calculated by compating to `BigFloat` calculations with the given

tpapp

######################################################################
# context: reading UInt8 lines from a gzipped stream
# - the real dataset has about 5e10 lines, this is a self-contaned MWE
# - the real dataset lines are then processed, this MEW is just about optimizing reading
# - line length can be bounded (relevant for buffered reading)
######################################################################

using CodecZlib

"""