Ident_men too long to be converted to long data for 2005

[thomasdouenne]

When trying to create a survey scenario with the 2005 survey, I get the following error:

OverflowError: Python int too large to convert to C long

Ident_men are of the order 10e11 for the 2005 survey. Do you know at which stage we should solve this issue? You can meet this error on almost any example, see for instance master_thesis/taux_effort_ticpe.py

[benjello]

What ident_men looks like ? Maybe you could modify it ?

[thomasdouenne]

No, I cannot find a good transformation, this is an example:

11012701001

But the last number could either be 0 or 2 so I see no general rule to simplify it.

[benjello]

Is changing values to 0 to n_men everywhere you need ident_men a possible strategy ?

On Mon, Sep 18, 2017 at 4:04 PM thomasdouenne notifications@github.com wrote:

No, I cannot find a good transformation, this is an example:

11012701001

But the last number could either be 0 or 2 so I see no general rule to simplify it.

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/openfisca/openfisca-france-indirect-taxation/issues/160#issuecomment-330232402, or mute the thread https://github.com/notifications/unsubscribe-auth/ABHxd-n29XnSJQNiIOM5yx-Qu8S16yU-ks5sjnhUgaJpZM4Pa8a2 .

-- Mahdi Ben Jelloul

[thomasdouenne]

Yes I can do that, but at which stage would you prefer? In the build_survey_data ?

[benjello]

Yes

[benjello]

@thomasdouenne : actually Ia am not sure. Maybe you should do it int the run_all (within openfisca-france-indirect-taxation)

[thomasdouenne]

Ok I can do this at that stage