# How to read a table from url
**URL:** https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985
**Category:** General Usage
**Tags:** dataframes, website, tables
**Created:** [November 29, 2022, 10:47am UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985 "2022-11-29T10:47:56Z")
**Posts on this page:** 9
**Page:** 1
### Author:  [@empet](https://discourse.julialang.org/u/empet)
#### Post date: [November 29, 2022, 10:47am UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/1 "2022-11-29T10:47:56Z")
With pandas I can read a table into a dataframe, as follows:
```julia
import pandas as pd
dfs = pd.read_html('http://stats.ioinformatics.org/countries/')#returns a list of tables
df=dfs[0]
df.head()#inspect header
```
I searched for a similar method in DataFrames.jl (more precisely in the forthcomming book, Julia for Data Analysis, v\_10, which I have pre-ordered), as well in the Julia for Data Science, but there is no approach or example of such a task. Following a suggestion given as answer to the same question, posted here, two years ago: [[Any equivalent to Pandas read\_html() in DataFrames.jl?](https://discourse.julialang.org/t/any-equivalent-to-pandas-read-html-in-dataframes-jl/55674))([Any equivalent to Pandas read\_html() in DataFrames.jl?](https://discourse.julialang.org/t/any-equivalent-to-pandas-read-html-in-dataframes-jl/55674)), I tried something like this:
```julia
using DataFrames, CSV, HTTP
read_remote_csv(url) = DataFrame(CSV.File(HTTP.get(url).body))
df = read_remote_csv("http://stats.ioinformatics.org/countries/")
```
but it displays the html contents, not a dataframe having as columns the table columns.
---
### Author:  [@nilshg](https://discourse.julialang.org/u/nilshg)
#### Post date: [November 29, 2022, 12:24pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/2 "2022-11-29T12:24:13Z")
My answer was for situations where the remote url returns a delimited file, which is what `CSV.File` parses. You are looking to extract a table from html, so need a library that parses html rather than delimited files, see:
> [@Scraping a html table from a website](https://discourse.julialang.org/t/scraping-a-html-table-from-a-website/47378):
>
> I am trying to obtain COVID-19 data from a website. The website has the data which I want but it is in a html table format I am looking for Julia tools to scrape the information from the html table. Something like this Is there an existing package in Julia for this? I am thinking about writing my own module to do this but I ask here first before writing my own module.
---
### Author:  [@empet](https://discourse.julialang.org/u/empet)
#### Post date: [November 29, 2022, 4:02pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/3 "2022-11-29T16:02:29Z")
Thank you for the link to Scraping a html table from a url. Unfortunately I have a low level of knowledge and skills in html/CSS. For the moment I will read the tables with `pd.read_html(url)`, save the corresponding dataframe as csv and re-read it in Julia.
---
### Author:  [@rafael.guerra](https://discourse.julialang.org/u/rafael.guerra)
#### Post date: [November 29, 2022, 6:57pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/4 "2022-11-29T18:57:52Z")
Check also this post on [TableScraper.jl](https://discourse.julialang.org/t/ann-tablescraper-jl-an-easy-way-to-scrape-well-formed-tables-from-webpages/61636)
---
### Author:  [@cormullion](https://discourse.julialang.org/u/cormullion)
#### Post date: [November 29, 2022, 7:11pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/5 "2022-11-29T19:11:08Z")
Beat me to it…
```julia
using TableScraper
using DataFrames
url = "https://stats.ioinformatics.org/countries/"
st = scrape_tables(url)
df = DataFrame(q=[],
Country=String[],
Host=String[],
G=[],
S=[],
B=[],
Total=[])
for row in first(st).rows
push!(df, row)
end
df.G = parse.(Int, df.G)
df.S = parse.(Int, df.S)
df.B = parse.(Int, df.B)
df.Total = parse.(Int, df.Total)
```
```julia
109×7 DataFrame
Row │ q Country Host G S B Total
│ Any String String Int64 Int64 Int64 Int64
─────┼─────────────────────────────────────────────────────
1 │ Albania 0 0 0 0
2 │ ? Argentina 1993 3 9 23 35
⋮ │ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
108 │ ? Yugoslavia 1 3 1 5
109 │ Zimbabwe 0 0 0 0
105 rows omitted
```
---
### Author:  [@empet](https://discourse.julialang.org/u/empet)
#### Post date: [November 30, 2022, 11:03am UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/6 "2022-11-30T11:03:44Z")
@cormullion
Thanks for your nice solution. I adopted it but replaced the last four lines of code by:
```julia
for name in names(df)[end-3:end]
df[!, name]= parse.(Int, df[!, name])
end
```
---
### Author:  [@rafael.guerra](https://discourse.julialang.org/u/rafael.guerra)
#### Post date: [November 30, 2022, 4:41pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/7 "2022-11-30T16:41:09Z")
Or without for loops:
```julia
using TableScraper, DataFrames
url = "https://stats.ioinformatics.org/countries/"
st = scrape_tables(url)
df = DataFrame(permutedims(reduce(hcat, first(st).rows)), [:q,:Country,:Host,:G,:S,:B,:Total])
df[!,[:G,:S,:B,:Total]] .= parse.(Int, df[!,[:G,:S,:B,:Total]])
```
---
### Author:  [@empet](https://discourse.julialang.org/u/empet)
#### Post date: [November 30, 2022, 5:31pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/8 "2022-11-30T17:31:00Z")
Yes, but the symbols :G, :S, :B, :Total, are repeated three times.
---
### Author:  [@rafael.guerra](https://discourse.julialang.org/u/rafael.guerra)
#### Post date: [November 30, 2022, 5:44pm UTC](https://discourse.julialang.org/t/how-to-read-a-table-from-url/90985/9 "2022-11-30T17:44:08Z")
Could be replaced by `4:7`, or assigned to a single variable, for instance.