I find Nushell even better for these usecases:

    $medias = open .\medias.csv

The above is the initial read and format into table.

I'm currently on my phone so can't go through all the examples, but knowing both PS and nu, nu has the better syntax.

EDIT:

Get data and view in table:

    let $medias = http get https://github.com/medialab/corpora/raw/master/polarisation/medias.csv
    $medias

Get headers:

    $medias | columns

Get count of rows:

   $medias | length

Get flattened, slight more convoluted (caveat there might be a better way):

    $medias | each {print $in}

Search rows:

    $medias | where $it.outreach == 'internationale'

Select columns:

    $medias | select foundation_year name

Sort file:

    $medias | select foundation_year name | sort-by foundation_year

Dedup based on column:

    $medias | uniq-by mediacloud_ids

Computing frequency and histogram

    $medias | histogram edito