remove unecessary coefficients from kw94 specifications

[peisenha]

• Date you used respy: 03/09/19
• respy version used, if any:
• Python version, if any:
• Operating System:

Description

I was simulating the example models kw94_ and noticed that there are several unnecessary covariates specified and their coefficients set to zero. Given the improved flexibility of the model specification, we can remove these now.

What I Did

params, options = rp.get_example_model("kw_94_one", with_data=False)
params

[tobiasraabe]

Yes, we are able to remove the coefficients, but at the same time the coefficients are present in the Monte Carlo estimation in Table 4. Thus, I would not remove them as we had to add them again for the validation exercises.

[peisenha]

We are discussing a different set of coefficients. I agree with you that we want to keep all coefficients that are estimated during the Monte Carlo exercise. But why do we need is_minor for example https://github.com/OpenSourceEconomics/respy/blob/develop/respy/tests/resources/kw_94_one.csv#L41 This only plays a role in KW1997.

…

On Tue, Sep 3, 2019 at 8:53 AM Tobias Raabe ***@***.***> wrote: Yes, we are able to remove the coefficients, but at the same time the coefficients are present in the Monte Carlo estimation in Table 4. Thus, I would not remove them as we had to add them again for the validation exercises. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#253?email_source=notifications&email_token=AA3QYJYTYL6MFB6SZOPB4Z3QHYCVLA5CNFSM4ITCROVKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5XGREQ#issuecomment-527329426>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA3QYJ3J3V6HW764EAXTJNLQHYCVLANCNFSM4ITCROVA> .

-- *Philipp Eisenhauer* Economist Mail eisenhauer@policy-lab.org Web www.eisenhauer.io Repository https://github.com/peisenha

[tobiasraabe]

I see. You are right.

[tobiasraabe]

Closed in #249.