define distributions directly in initial parameters for models

sbenthall commented 4 years ago

Related to #227

Currently, when the parameters of distributions are given to models, they are listed as unstructured fields whose meaning depends on their name, e.g.: aNrmInitMean. aNrmInitStd. pLvlInitMean. pLvlInitStd, ...

It can be changed so that these parameters look like: aNrmInit : LogNormal(0, 1).

sbenthall commented 3 years ago

This is related to #664 , because these distributions are often '"time-varying".

It looks like the expectation is that TranShkStd is a list even when there's a single value and so is "time-varying" even though it is static across time.

https://github.com/econ-ark/HARK/blob/master/HARK/ConsumptionSaving/ConsIndShockModel.py#L2020-L2023

I don't see any reason for this except for haste in the early coding.

sbenthall commented 3 years ago

This feature would make HARK closer to Dolo. Dolo requires users to specify the distribution or process driving values of exogenous shocks in configuration. E.g. https://github.com/EconForge/dolo.py/blob/master/examples/models_/rbc.yaml#L72-L73

llorracc commented 3 years ago

No, this is deliberate.

Remember that IndShockConsumerType is used for finite horizon (life cycle) models as well as infinite horizon models.

There is a lot of variation by age in the magnitude of income shocks, both transitory and permanent. Actually, that age-related variation can have important implications for things like the profile of stockholding patterns by age. So it is deliberate that the default is to allow TranShkStd and PermShkStd to vary with time/age.

On 2021-04-04, at 13:52, Sebastian Benthall @.***> wrote:

This is related to #664 , because these distributions are often '"time-varying".

It looks like the expectation is that TranShkStd is a list even when there's a single value and so is "time-varying" even though it is static across time.

https://github.com/econ-ark/HARK/blob/master/HARK/ConsumptionSaving/ConsIndShockModel.py#L2020-L2023

I don't see any reason for this except for haste in the early coding.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

sbenthall commented 3 years ago

I see now I've conflated two things in this ticket.

The original ticket is specifically for the Initial conditions for an agent. It's "birth values". These are not time-varying.
Yes, of course, allowing time and age-varying income is an important feature and deliberate.
My comment here was that it looks a little funny to use a list for a singleton, non-time-varying value. I'm afraid I've muddied the water by bringing it into this ticket. This design choice seems to be inconsistent across the library.
My recommendation, based on the design goal to get things closer to Dolo, is to create a Process class for time-varying exogenous values.

So, rather than giving a model:

init_cyclical['PermShkStd'] = [0.1, 0.1, 0.1, 0.1]
init_cyclical['TranShkStd'] = [0.1, 0.1, 0.1, 0.1]

as part of its initial parameters, you would give it:

init_cyclical['exogenous']['PermShk'] = CyclicProcess(LogNormal, mu=[1,1,1,1], std=[0.1,0.1,0.1,0.1])
init_cyclical['exogenous']['TranShk'] = CyclicProcess(LogNormal, mu=[1,1,1,1], std=[0.1,0.1,0.1,0.1])

The advantage of this is then get_shocks would have everything it needs to sample the shocks without any model-specific code.

If the distribution was not time-varying, I'd suggest:

init['exogenous']['PermShk'] = Sample(LogNormal, mu=1, std=0.1)
init['exogenous']['TranShk'] = Sample(LogNormal, mu=1, std=0.1)

and a related configuration for an AgeProcess.

It would be a little bit harder, but not impossible, to allow these Process definitions to refer to the existing parameters TransShkStd, etc. That might be better for backward compatibility.

llorracc commented 3 years ago

OK, that post clarifies things (and I've read further back in the thread as well, so now am more up to speed on the issues here).

What's not clear to me is whether it makes sense to tackle this set of issues in the specific context of shocks now, because we are facing exactly the same issue with every other aspect of the problem. Our discussion of how to create an architecture in which we allow for successive "stages" of the Bellman problem is essentially saying that not only do we need to allow shock processes to change (arbitrarily) from one stage to the next, we need to allow every single aspect of the problem except the value function to change: State variables, control variables, optimality conditions, solvers, etc. If we had an architecture that worked for all of those elements, it would also work for distributions.

sbenthall commented 3 years ago

I see the "stage" issue as quite different. See our terminology chart.

Exogenous shocks are occurring once per period. There may be multiple stages per period (one per control variable), but only one instance of the shock variable in the period.

I'm making these proposals because they would make HARK look more like Dolo.

I believe the question of how Dolo deals with cases where there is more than one control variable is unresolved, whereas Dolo handles shocks more elegantly than HARK currently does. So I see the questions as mostly orthogonal.

llorracc commented 3 years ago

But, per our earlier discussion that the "problem" may change, say, at different ages of life, if we can have an architecture can handle that for all the other things that are MORE complicated than the distributions (because, for example, there may be several transition equations or whatever inside a period, but only one draw of exogenous shocks), then we will surely also want to use the same architecture for keeping track of the shocks as for keeping track of everything else about the problem.

Further, I do not see any technical reason to structure things in such a way that there can only be one draw of shocks per period. Maybe a good technical case for this exists and I'm missing it, but if, for example, one of the decisions that a person makes in the period is whether or not to take some risk, or which of several kinds of risks to take, it would be wasteful to precompute shock distributions for all possible choices if the person might not actually make them.

Also, the ability to deal with these issues is one of the few ways in Dolo is going to change to achieve HARK-like capabilities, instead of the other way around.

I'm sure that in the revamping Pablo will keep his treatment of distributions as processes, so maybe you're right that we should go ahead and emulate that now. But to the extent that we will need to re-engineer that after we get the whole stages thing worked out, it might be more efficient at least to iron out our plans for the stages thing before tackling the distributions.

sbenthall commented 3 years ago

I'm sure that in the revamping Pablo will keep his treatment of distributions as processes, so maybe you're right that we should go ahead and emulate that now.

I will take this as a concession of the point with respect to the narrow scope of this issue, which I see as being strictly about how to configure the initial distributions and exogenous shocks of HARK models that are currently in the library.

I do have thoughts about the other points you're bringing up and look forward to further conversations to iron out the plans for stages.

sbenthall commented 2 years ago

See #1024 for steps towards this.

sbenthall commented 2 years ago

See this note about correlated shocks: https://github.com/econ-ark/HARK/issues/1106#issuecomment-1025908532

Mv77 commented 2 years ago

The work I did on this in https://github.com/econ-ark/HARK/pull/1121 was blocked by the simulation test issue.

sbenthall commented 2 years ago

Shoot, ok. blocked by #1105

mnwhite commented 2 years ago

I looked at the linked example, and I don't think there's an inconsistency there. In the basic ConsIndShock model, Rfree is a single float because it's assumed to be exogenous to the individual, and there's only one asset to hold. Likewise, the coefficient of relative risk aversion is constant because we assumed the agent has the same intraperiod preferences in all periods. It works out that changing Rfree to being time-varying (and changing the input to a list) will result in the correct solution (but the simulation code would need to be adjusted), while trying to set CRRA to be time-varying would result in an incorrect solution because the solution code makes assumptions about the structure of the problem, which would be violated if CRRAt \neq CRRA{t+1}. The solver would run, but the produced consumption function would be somewhat wrong.

Similarly, RiskyDstn is time-invariant because it describes the returns on a second asset, which is external to the agent (as a matter of economic assumption). And just like Rfree, changing it to be time-varying will result in the correct solution being produced by the code, but the simulation code would need updating. There's no inconsistency here between the form of the model inputs and the form of the model output. It was a deliberate design choice, not done out of haste or laziness.

As for raw parameters versus objects in dictionaries, it's somewhat a matter of taste. The design intent was twofold. First, to be able to have all model parameters sit in a JSON file, which could be read in by HARK or Mathematica or Matlab or whatever to be able to easily compare solution methods/libraries, working from the same parameter file. Second, to make model estimation (or other usage) code easier / more intuitive to write. Suppose you want to see how some model outcome of interest varies with a distribution parameter, in a simple setting with just one AgentType instance. The code for that looks like this:

ParamVec = some vector OutVec = np.zeros_like(ParamVec) MyType = AgentTypeSubclass(**baseline_dictionary) for j in range(ParamVec.size): ....MyType(param_name = ParamVec[j]) ....MyType.update() ....MyType.solve() ....MyType.initialize_sim() ....MyType.simulate() ....OutVec[j] = extract_interesting_information(MyType.history)

plt.plot(ParamVec, OutVec) plt.xlabel(param_name) plt.ylabel(interesting statistic name) plt.show()

Estimation code looks similar: just cross off the parameter initialization and plotting code, have more parameters get set, and throw an off-the-shelf optimizer around the whole thing.

My intent was that if someone wanted an alternate parametric specification for some distribution, they'd just make a subclass of the AgentType class and overwrite its updateThatDstn() method.

On Wed, May 11, 2022 at 11:40 AM Sebastian Benthall < @.***> wrote:

Shoot, ok. blocked by #1105 https://github.com/econ-ark/HARK/issues/1105

— Reply to this email directly, view it on GitHub https://github.com/econ-ark/HARK/issues/620#issuecomment-1123943607, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADKRAFNFG2GOQVY52TZEOV3VJPIF7ANCNFSM4MEZ7E4Q . You are receiving this because you are subscribed to this thread.Message ID: @.***>

llorracc commented 2 years ago

"trying to set CRRA to be time-varying would result in an incorrect solution because the solution code makes assumptions about the structure of the problem, which would be violated if CRRAt \neq CRRA{t+1}. "

Yes. Things like this are one of the key reasons I want to move to a structure where everything the agent at t knows about the solution at t+1 is explicitly passed as part of the structure of an explicit input to the time-t agent.

Several steps I want to take will be helpful here:

Move realization of all shocks to the instant after the beginning of the period (before any actions have been taken).
Construct our expectations objects (e.g., my former "gothic v") as of the beginning of the period.
Give a unique and explicit name to each state variable as of the beginning of the period (like, my k{t} is the previous period's a{t-1} -- before the agent knows realizations of the stochastic shocks that ultimately yield my b_{t}

In principle, that would allow us to have any sequence of values of relative risk aversion. Their consequences would be entirely embodied in the value, marginal value, etc functions at the beginning of the period, so the t-1 agent would only need to know the t "expected" value function as a function of the beginning-of-period states (k).

RiskyDstn is time-invariant because it describes the returns on a second asset, which is external to the agent (as a matter of economic assumption).

But at present if we want to handle perfect foresight MIT shocks to the aggregate (like, people might worry that the rate of return is riskier in election years, or in the millenium year, or whatever), we have to do that by having the sequence of distributions laid out as being different by age, and use the fact that if a person is one year older it is because they have moved one year into the future in calendar time. This example may seem a bit artificial being able to handle it is core to being able to use our toolkit to generically handle a vast class of macro models -- with perfect foresight MIT shocks.

sbenthall commented 5 months ago

@mnwhite the original scope of this issue was to define shock distributions directly for models, to make them more explicit.

This is happening in the new model format, here: https://github.com/econ-ark/HARK/blob/master/HARK/models/consumer.py

I'm satisfied by this.

I know you've been working on the older models, with the 'constructions' workflow, which I don't think is directly related to this.

Do you see any reason not to close this issue as completed? I think it can be closed.

mnwhite commented 5 months ago

Close it.

On Mon, Jun 3, 2024 at 3:27 PM Sebastian Benthall @.***> wrote:

@mnwhite https://github.com/mnwhite the original scope of this issue was to define shock distributions directly for models, to make them more explicit.

This is happening in the new model format, here: https://github.com/econ-ark/HARK/blob/master/HARK/models/consumer.py

I'm satisfied by this.

I know you've been working on the older models, with the 'constructions' workflow, which I don't think is directly related to this.

Do you see any reason not to close this issue as completed? I think it can be closed.

— Reply to this email directly, view it on GitHub https://github.com/econ-ark/HARK/issues/620#issuecomment-2145958566, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADKRAFKB7TRFOGXTYPKEFJ3ZFS7SNAVCNFSM4MEZ7E42U5DIOJSWCZC7NNSXTN2JONZXKZKDN5WW2ZLOOQ5TEMJUGU4TKOBVGY3A . You are receiving this because you were mentioned.Message ID: @.***>

econ-ark / HARK

define distributions directly in initial parameters for models #620