Loading Fannie Mae takes unusually long

[xiaodaigh]

Update

On Julia 1.2 I get the below
I have included a MWE that can actually run by downloading the Fannie data direct from https://docs.rapids.ai/datasets/mortgage-data

However, loading the data takes a really long time! I waited for 2 hours and it didn't finish.

On Julia 1.1.1 I get the below

┌ Warning: In c:/data/perf-jld/Performance_2000Q2.txt line 2063379 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q1.txt line 2162075 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q2.txt line 2063402 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q1.txt line 2162113 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q2.txt line 2063482 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q1.txt line 2162161 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q2.txt line 2063515 has 29 fields but 31 fields are expected. Skipping row.
└ @ TextParse C:\Users\RTX2080\.julia\packages\TextParse\tFXtC\src\csv.jl:372
┌ Warning: In c:/data/perf-jld/Performance_2000Q1.txt line 2162202 has 29 fields but 31 fields are expected. Skipping row.

Code

using Distributed, Statistics
#addprocs(4)

@time @everywhere using JuliaDB, Dagger

##############################################################
# Download & Extract data
###############################################################

;wget http://rapidsai-data.s3-website.us-east-2.amazonaws.com/notebook-mortgage-data/mortgage_2000.tgz
;tar xzvf mortgage_2000.tgz


##############################################################
# Specify the types of columns
###############################################################

const fmtypes = [
    Int64,                     String,     Union{String, Missing},     Union{Float64, Missing},    Union{Float64, Missing},
    Union{Float64, Missing},    Union{Float64, Missing},    Union{Float64, Missing},    Union{String, Missing},     Union{String, Missing},
    Union{String, Missing},     Union{String, Missing},     Union{String, Missing},     Union{String, Missing},     Union{String, Missing},
    Union{String, Missing},     Union{String, Missing},     Union{Float64, Missing},    Union{Float64, Missing},    Union{Float64, Missing},
    Union{Float64, Missing},    Union{Float64, Missing},    Union{Float64, Missing},    Union{Float64, Missing},    Union{Float64, Missing},
    Union{Float64, Missing},    Union{Float64, Missing},    Union{Float64, Missing},    Union{String, Missing},     Union{Float64, Missing},
    Union{String, Missing}]

datapath = "perf/"
ifiles = joinpath.(datapath, readdir(datapath))


colnames = ["loan_id", "monthly_rpt_prd", "servicer_name", "last_rt", "last_upb", "loan_age",
    "months_to_legal_mat" , "adj_month_to_mat", "maturity_date", "msa", "delq_status",
    "mod_flag", "zero_bal_code", "zb_dte", "lpi_dte", "fcc_dte","disp_dt", "fcc_cost",
    "pp_cost", "ar_cost", "ie_cost", "tax_cost", "ns_procs", "ce_procs", "rmw_procs",
    "o_procs", "non_int_upb", "prin_forg_upb_fhfa", "repch_flag", "prin_forg_upb_oth",
    "transfer_flg"];

@time jll = loadtable(
    ifiles,
    output = "data/fm.jldb/",
    delim='|',
    header_exists=false,
    filenamecol = "filename",
    #chunks = length(ifiles),
    #type_detect_rows = 20_000,
    colnames = colnames,
    colparsers = fmtypes,
    indexcols=["loan_id", "monthly_rpt_prd"]);