julia_coursera/COVID_data_raw.ipynb

409 lines
15 KiB
Plaintext
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"using DelimitedFiles"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"#? readdlm"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"269×117 Array{Any,2}:\n",
" \"\\ufeffProvince/States\" \"Country/Region\" … \"5/13/2020\"\n",
" \"Confirmed\" \"Globally\" 4170424 \n",
" \"Deaths\" \"Globally\" 287399 \n",
" \"Confirmed\" \"China\" 84458 \n",
" \"Confirmed\" \"Outside of China\" 4085966 \n",
" \"Deaths\" \"China\" … 4644 \n",
" \"Hubei \" \"China\" \"\" \n",
" \"Guangdong\" \"China\" \"\" \n",
" \"Henan\" \"China\" \"\" \n",
" \"Zhejiang\" \"China\" \"\" \n",
" \"Hunan\" \"China\" … \"\" \n",
" \"Anhui\" \"China\" \"\" \n",
" \"Jiangxi\" \"China\" \"\" \n",
" ⋮ ⋱ \n",
" \"Confirmed\" \"\" 163201 \n",
" \"Deaths\" \"\" 6578 \n",
" \"Confirmed\" \"\" 1780316 \n",
" \"Deaths\" \"\" … 159799 \n",
" \"Confirmed\" \"\" 110932 \n",
" \"Deaths\" \"\" 3746 \n",
" \"Confirmed\" \"\" 284270 \n",
" \"Deaths\" \"\" 9259 \n",
" \"Confirmed\" \"\" … 1781564 \n",
" \"Deaths\" \"\" 106504 \n",
" \"Confirmed\" \"\" 49429 \n",
" \"Deaths\" \"\" 1500 "
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"coviddata = DelimitedFiles.readdlm(\"who_covid_19_sit_rep_time_series.csv\",',',header=false)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"269×2 Array{Any,2}:\n",
" \"\\ufeffProvince/States\" \"Country/Region\" \n",
" \"Confirmed\" \"Globally\" \n",
" \"Deaths\" \"Globally\" \n",
" \"Confirmed\" \"China\" \n",
" \"Confirmed\" \"Outside of China\"\n",
" \"Deaths\" \"China\" \n",
" \"Hubei \" \"China\" \n",
" \"Guangdong\" \"China\" \n",
" \"Henan\" \"China\" \n",
" \"Zhejiang\" \"China\" \n",
" \"Hunan\" \"China\" \n",
" \"Anhui\" \"China\" \n",
" \"Jiangxi\" \"China\" \n",
" ⋮ \n",
" \"Confirmed\" \"\" \n",
" \"Deaths\" \"\" \n",
" \"Confirmed\" \"\" \n",
" \"Deaths\" \"\" \n",
" \"Confirmed\" \"\" \n",
" \"Deaths\" \"\" \n",
" \"Confirmed\" \"\" \n",
" \"Deaths\" \"\" \n",
" \"Confirmed\" \"\" \n",
" \"Deaths\" \"\" \n",
" \"Confirmed\" \"\" \n",
" \"Deaths\" \"\" "
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"first2 = coviddata[:,1:2]"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"269×114 Array{Any,2}:\n",
" \"1/21/2020\" \"1/22/2020\" \"1/23/2020\" … \"5/13/2020\"\n",
" 282 314 581 4170424 \n",
" \"\" \"\" \"\" 287399 \n",
" 278 309 571 84458 \n",
" 4 5 10 4085966 \n",
" \"\" \"\" \"\" … 4644 \n",
" 258 270 375 \"\" \n",
" 14 17 26 \"\" \n",
" \"\" 1 1 \"\" \n",
" \"\" 5 5 \"\" \n",
" \"\" 1 1 … \"\" \n",
" \"\" \"\" \"\" \"\" \n",
" \"\" 1 2 \"\" \n",
" ⋮ ⋱ \n",
" \"\" \"\" \"\" 163201 \n",
" \"\" \"\" \"\" 6578 \n",
" \"\" \"\" \"\" 1780316 \n",
" \"\" \"\" \"\" … 159799 \n",
" \"\" \"\" \"\" 110932 \n",
" \"\" \"\" \"\" 3746 \n",
" \"\" \"\" \"\" 284270 \n",
" \"\" \"\" \"\" 9259 \n",
" \"\" \"\" \"\" … 1781564 \n",
" \"\" \"\" \"\" 106504 \n",
" \"\" \"\" \"\" 49429 \n",
" \"\" \"\" \"\" 1500 "
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"restofarray = coviddata[:,4:end]"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"269×116 Array{Any,2}:\n",
" \"\\ufeffProvince/States\" \"Country/Region\" … \"5/13/2020\"\n",
" \"Confirmed\" \"Globally\" 4170424 \n",
" \"Deaths\" \"Globally\" 287399 \n",
" \"Confirmed\" \"China\" 84458 \n",
" \"Confirmed\" \"Outside of China\" 4085966 \n",
" \"Deaths\" \"China\" … 4644 \n",
" \"Hubei \" \"China\" \"\" \n",
" \"Guangdong\" \"China\" \"\" \n",
" \"Henan\" \"China\" \"\" \n",
" \"Zhejiang\" \"China\" \"\" \n",
" \"Hunan\" \"China\" … \"\" \n",
" \"Anhui\" \"China\" \"\" \n",
" \"Jiangxi\" \"China\" \"\" \n",
" ⋮ ⋱ ⋮ \n",
" \"Confirmed\" \"\" 163201 \n",
" \"Deaths\" \"\" 6578 \n",
" \"Confirmed\" \"\" 1780316 \n",
" \"Deaths\" \"\" … 159799 \n",
" \"Confirmed\" \"\" 110932 \n",
" \"Deaths\" \"\" 3746 \n",
" \"Confirmed\" \"\" 284270 \n",
" \"Deaths\" \"\" 9259 \n",
" \"Confirmed\" \"\" … 1781564 \n",
" \"Deaths\" \"\" 106504 \n",
" \"Confirmed\" \"\" 49429 \n",
" \"Deaths\" \"\" 1500 "
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"coviddata = [first2 restofarray]"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(269, 116)"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"size(coviddata)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"213×115 Array{Any,2}:\n",
" \"Country/Region\" \"1/21/2020\" … \"05/12/2020\" \"5/13/2020\"\n",
" \"China\" 278 84451 84458 \n",
" \"Singapore\" 0 23822 24671 \n",
" \"Japan\" 1 15874 16024 \n",
" \"Republic of Korea\" 1 10936 10962 \n",
" \"Philippines\" 0 … 11086 11350 \n",
" \"Australia\" 0 6948 6964 \n",
" \"Malaysia\" 0 6726 6742 \n",
" \"New Zealand\" 0 1147 1147 \n",
" \"Vietnam\" 0 288 288 \n",
" \"Brunei Darussalam\" 0 … 141 141 \n",
" \"Cambodia\" 0 122 122 \n",
" \"Mongolia\" 0 42 42 \n",
" ⋮ ⋱ \n",
" \"South Sudan\" 0 174 174 \n",
" \"Malawi\" 0 57 58 \n",
" \"Eritrea\" 0 39 39 \n",
" \"Angola\" 0 45 45 \n",
" \"Zimbabwe\" 0 … 36 37 \n",
" \"Botswana\" 0 24 24 \n",
" \"Burundi\" 0 19 27 \n",
" \"Gambia\" 0 22 23 \n",
" \"Namibia\" 0 16 16 \n",
" \"Seychelles\" 0 … 11 11 \n",
" \"Mauritania\" 0 8 8 \n",
" \"Comoros\" 0 11 11 "
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"coviddatabycountry = coviddata[42:254,2:end]"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(213, 115)"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"size(coviddatabycountry)\n",
"rows,cols = size(coviddatabycountry)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Array{Any,2}"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"typeof(coviddatabycountry)"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
"using Dates"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"2020-01-02T00:00:00"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Dates.DateTime(\"1/2/2020\",\"m/d/y\")"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"covidbycountry = Array{Any}(undef,cols,rows);"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"for j = 1:cols\n",
" for i = 1:rows\n",
" covidbycountry[j,i] = coviddatabycountry[i,j]\n",
" end\n",
"end"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"\"Date\""
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"covidbycountry[1,1] = \"Date\""
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"DelimitedFiles.writedlm(\"who_covid_trans.csv\",covidbycountry,',')"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Julia 1.2.0",
"language": "julia",
"name": "julia-1.2"
},
"language_info": {
"file_extension": ".jl",
"mimetype": "application/julia",
"name": "julia",
"version": "1.2.0"
}
},
"nbformat": 4,
"nbformat_minor": 4
}