today passed as:
cleaning the home as i wake up and had breakfast. (since mom had suggested to tidy up textbooks cupboards etc and untidiness around as they saw home last night. )
i made as food pasta + meatballs. and i would put some to freezer since it has been a lot.
i ate too much food.
i had bath after cleaning home.
i over talked with mom and dad :D i call them a lot today:D (first to show cleaned home, second to show food i baked)
now some coffee and simulation study time. actually its always in my mind some part of it. yupp studying. starts now.
i think i over ate the food i made:S it were my second meatballs cooking try in my life the first one failed a lot in terms of taste. this one were better. when i make it like just pressed meat on oven on oven paper it becomes too much. i thing meatballs makes to consume less cause when i consume too much i feel heavy all entire day afterwards. this time even if i feel is m very full of food i think i hadn't overeaten unhealthily i think pasta makes me feel overeaten. when you overeat meat like foods it feels so heavy and is surely unhealthy. so i think even if is not easy to cook, i would prepare meat food as meatballs since pressed all in oven becomes worse version of food. but i hated this constant routine of preparing food. i wish there were food subsitutes and i don't have to cook food. it just takes lots of energy to cook food. i figured out meat never becomes tasty if i don't make it like this in today's version. ok. i wish making food were easier. its very hard. it took 1 hour like duration to prepare pasta meatballs :S but i have to get used to. otherwside i have to constantly order which wouldn't be any cheap if everytime i order food. maybe i should investigate some cooking devices to cook food easier. in anyway its very hard to cook prepare food constantly. but have to surely. some side hassle of life. which is constant. cause we always have to eat daily. i don't know how to optimize this making food process of my life. its an activity i don't quite like at all. i wish i were like people who likes to cook food. i definitely don't any like. its my second meatballs cooking in my life. this time didn't got bad like first version. i previously also tried just pressing entire minced meat to oven plate so i don't have to work alot to prepare but that becomes very untasty afterwards. ok maybe i find some frozen versions ready version of these foods aha. so that i don't have to spend 1 hour to cook it:S its just so uninteresting to cook food:S (i wish i liked cooking food).
ok back to studying in 10 minutes. first day of holiday and simulation project's active work days started yuppp:) i would have 10 or 15 minutes more laziness then would start work day (for simulator project).
ok surely i would also help people whom has problems in finding food when i have economic capabilities. when everytime i talk about food i prepare, people whom lacks food comes to my mind making me feel kind of bitter kind of uncomfortable about it. and also there happens food waste alot i heard some startups hassling that to reduce it. anyway everytime i tell about food i prepared, i feel uncomfortable for people lacking access to nutritions. if i have someday after project some big economic capabilities i would surely workon somehow on this topic on some part of it. since it is very bad that nutrition topics don't reach everyone and it is very sad topic. and i would surely work on to fix it when i have my economic capabilities to work on it. maybe molecule engineering could be also done to build food synthetically. so that food becomes a very cheap resource and everyone has access to all nutrititons easily. ok lets note down this idea. i liked it. i just see there is imbalance in everyplace's economy. and it makes food access unbalanced. and it feels totally sad due to that. ok i think molecule engineering might also work to fix this issues to make food named topic cheapest as possible with making most expensive type foods in every economy cheapest with these synthetic options. yupp. i think i would surely work on to make food cheaper than anything in life. and i would also want to read some economics to understand why pricing differs so much about food in every economy. for instance i think in Germany meat is quite cheap. but in Turkey it were very expensive. beiong engineer i could take. i don't know any about economy so i don't understand pricing differences of core food products in every each country. actually people also in Turkey eat lots of meat even if is expensive. but the pricing difference took my attention of such products in each different country. i want to study economics to understand why. but other else on each part of world the food access capabilities is not same i think. and this is very sad. i wish to work on synthetic food production to make food named topic cheap everywhere on world. yupp i think molecule simulator is also good for that. since we could simulate emulate biological constructs to investigate how to synthethize for instance cow meat like stuff. based on its biological constituents. producing synthetic meat has been a hot topic to startstups as i read about. but i also want to work on these on molecular engineering level. after my molecule simulator gets ready. yupp second project's topic has been selected like this upon wind turbine designs gets completed. or maybe i wont work but would setup a team to work on it based on simulator i created. so that i direct the project but not invest that much time. but surely this would be second project. if i create lots of capital from windturbine i think i would be manage to setup a team for this second project. goal: making food cheapest resource in world in everywhere not only in strong economies. ok.working for against inequality issues in things like access to food. i selected second project's topic as. and goal would be narrative would be: food should be the cheapest resource in world. (in everywhere).
-----
ok lets not mention even.
------
ok i blogged alot again with spending lots of time to blogging:S
ok back to simulation study time in 15 minutes. which i think i would build my first millions in this holiday:D yayy:D i definitely thnk i would really revolutionize wind turbine designs. this holiday. soon :) 2 weeks later i think:)
just one paragraph of blog more: i had to stop studying to simulator every start of week. every weekend i would study then i would stop by week comes. this time i have also week time to simulators study:) its very exciting i have 3.5 weeks to finish my mvp:D its so happy:) never had this much time to this project before, i always had weekends and always happent context switches like you study some topic then stop!!! week starts, i focus on work tasks. then weekend comes then again context switch starts to this project and stops agaiin.. this time i have all days for simulator since i am on holiday:) this is so happy since i really think i have alot time to build my mvp and my mvp would be ready:D yayy:D
i would have just 20 minutes of more laziness then start my work day for simulator project:) to build first mvp:D
i think soon i would live in this type flat: (guh's flat goals be as) :
https://www.youtube.com/watch?v=oakA7RLvmWs&t=1177s
(my lovely future living room be this as) :)
ok i would have just 15 minutes some fun activity (i like solitaire with cards and since i cleaned my place to sit and open solitaire (it were filled with books/papers until mom suggested to clean) now i could have some ssolitaire fun. i should learn other types of solitaire. or such stuff. ok some emptly passing time period with solitaire which seems fun:) (thanks to mom whom suggested to clean those parts of home:D) now i can resume solitaire fun sometimes.
coming to fixing verlet integrator studies part:
thinking of the curve we try to differentiate/integrate (r(t)):
yuppp i think r n times differentiated and r n+2 times differentiated goes in 3 scenarios: at that time instant:
either the series goes as: -1 over(m) *d(n+2m) applied to r = d(n) applied to r (for any m) (this is surely true for higher orders of m when any order is equal higher orders are also equal) changes
or. d(n+2m) applied to r > d(n) applied to r (for any m) (currently going to be thought of for other values of m)
or d(n+2m) applied to r < d(n) applied to r (for any m)(currently going to be thought of for other values of m)
(or not for any m i would double check this now. ) i try to figure out possible scenarios of the curve that we wish to integrate. i mean its high degree derivatives at single point. i need to double check what i written above. it might be wrong. i would write correct version.
(for 8 possible subspace directions these curves might go to. for each of them might have these 3 states. but i need to check frenet serret stuff to get to know more since i actually thought in just a plane but without the B part of that r(t) curve)
ok would investigate the dynamics of d(n)(r(t)) (differentiate r(t) for n times) and d(n+2m)(r(t)) for each case than i would have idea of the remainder part of taylor expansion. the applications says that the remainder part goes to 0 as t goes to infinity since otherelse the new r(t) would have infinite positions. ok i think i might check measure theory sides at some parts of this making verlet integrator better study or might not need to. lets see. ok this part of the task also looks quite fun. since started to include differential geom stuff. anytopic with diff geometry stuff is fun. though i just saw dif. geometry's only introduction parts yet in just calculus texts. i know nothing yet from diff geom. anyway. continuing to study to task.
----
so that; we just need to infer the propagation with respect to factorial's rate of change when assessing impact of remaining part in 2.d case.
in first case, it would be investigated. it might be even infinity value :D in some config. anyway. would investigate each cases case by case.
in second case, we need to figure out the resulting value
in third case, it would be even much smaller value than the first case
delta t conf also i investigate how to setup. yupp i resumed verlet integration studies to include those remainder parts impacts in that r(t+delta_t) + r(t-delta_t) version.
since last week i were ill i couldn't finish this part of trying to fix verlet integration's some method of not covering remaining part of taylor expension which seems wrong/incorrect to me.
i am so eager to move to akka part since akka is severe levels of fun and coding in akka is severely fun (though i would need to revise scala little bit from a language cheat sheet page) . i want to fastly move to coding of simulator underlying structure but i need to solve this integrator part to try to make it more correct. i think little effort is left on this part.
wov:D its not hard to write a molecule simulator. or i might be thinking differently when i come to torque or multi centered molecules and also dihedral structures with angles, all needs to be simulated. ok. but still after those also gets accomplished, it seems as ok i think i would have my wind turbine product soon :D(and have my first millions in my life:D) (ok in 2 weeks i guess or almost 3 weeks it would be ready).:D
good to know: came up to gamma function: (for checking differentiation of factorial)
https://math.stackexchange.com/questions/300526/derivative-of-a-factorial#:~:text=The%20derivative%20of%20a%20function,the%20derivative%20at%20integer%20values.
i wish i knew maths like these discussions. i don't know any yet.
ok would learn about these to know more: https://en.wikipedia.org/wiki/Harmonic_number
but later.. lets fix this integrator first. but maybe i would have to use these stuff from harmonic number and digamma function like stuff. anyway. first check the possible scenarios for different values of m first them move to that infinite series of (dn(r(t)) /n!) * (delta_t over n) for n from 3 to infinity. ok.
ok i want to move to simulator but this part i should solve and design a better verlet version first. cause verlet until investigated thoroughly does makes no sense to me in its current design version it seems incorrect (since is not complete). but i don't know if i could make it better. but would try. and currently working on. bye to blogging for a while again. i am eagerly awaiting time to move to the simulator part and akka part:) i should fastly finish this part asap so that i can code with akka which is severe levels of fun as far as i remember.
ok i revised dif geometry topic to double check the parameterization to see it were arc-length based. but still there is conversion chances available i think and i think arclength based world is easier to model possible scenarios cases. for how the unknown curve iterates that we don't correctly model in verlet with having the new position be more with the remainder error term. even if that remainder is dimunitive number that doesn't mean we can just ignore that part like verlet algo does.
ok moving to arclength parameterization to simplify various curve models to investigate how d(n+2*m) for m 0 to infinity behaves. i think we could find out some thing that is inductive or some formulation that enables us to model the remainder error part of taylor polynomial with more models instead of just a simple errorpart's exponentiation methodology which they named as Verlet that does not travel tons of configurations phases of the molecular dynamics model. we need to separate cases of to see how we could eliminate remainder safely in which possible curvature models for the expected curve model, or we could have a multi way of iterating the model with different outcomes which means exponentially increasing different iteration configurations of the system which might actually have an expected nonreversiblity definition which would be aligned with entropy definitions i guess. i feel as there must be at least 2 such possible iterations at the same time which is aligned with entropy def that we don't have a reversible system as expected by entropy definitions. but if it turns out single possible curve model for iteration of system than entropy principle is not exactly correct but just a side effect but there actually exists revertiblity.
ok so i need to tackle this problem through possible curve models and their conversions in usual time based representation to try to find out some inductive rule set for d(n+2*m) differentiations. i belive there might be found out some rule systems in between d(n) and d(n+2*m) iterations that enables more correct modeling. current model exponentiates errors(Verlet one). we need to change this. as tokd i think arc length based curve abstractions might figure out to abstract possible cases and figure out some inductive mechanism in d(n+2*m) differentiations if we took Taylor to more terms taking error terms impact much more lesser in some situations according to possible behavior sets of d(n +2m)
i think we could call verlet an error exponentiator :P but if i cant find any possible solutions i might also need to be bound to verlet either. or again would try to modify it with some extra steps to take back that error terms impact from iteration.
i just still cant believe verlet is accepted in its current form. maybe people analyzed what i am analyzing right now and later decided to use an erroneous form since there is no solution. this might be why verlet is like this. other else it seems so erroneous. how could you not take care of error term of taylor polynomial? i don't understand. how they use verlet like that i just don't yet understand.
ok time for arc-lenght <-> time based parameterizations conversions time and modeling curves simpler in arc length side. ok creating a clustering for possible curve iterations regarding configurations of this problem without using verlet named error exponentiator named stuff i don't any understand how people use yet.
ok lets note down what i thought to continue on tomorrow morning. model in arclenght possible curve models and investigate possible time based model conversions possible inductions on dns with d(n+2m)s. goal is to find inductions on taylors long list or short list related to findings on where we could have E(n+1) side pause the list. some acceleration inductive analysis on possible curve models time but not today. i postpone this to tomorrow. since i woke up early today. ok since this part is an important part of problem, i would spare entire tomorrow to also possible better than verlet solutions investigations until end of tomorrow. i am pretty sure i would find some inductions or results that would give insight on how where to stop with E(n+1) term. i mean i think i would figure out some results by tomorrow. then i wish i could start coding this integrator part of simulation on Monday. ok i am not stressing myself to finish it by noon or by evening, i give endof tomorrow as deadline to finish this task to myself. if i had infinite period i wont either give any deadlines but i have just only 3.5 weeks holiday so i have no buffers to be any slow. i need to have a product by 3.5 weeks atmost. so by end of tomorrow somehow someway verlet investigation should finalize with some beneficial results. i am hopeful that i would accomplish that important subtask of this project. ok. now resting time. ok at most i could give 1 more day to verlet side. other else latest on Tuesday morning i would start coding simulation. ok. but i wish i accomplish on tomorrow. now rest time. since i woke up very early.
initial indctn i mntd were about arc length based thinking curves. but thinkng nrml acclrtn eithr has some harmonic bhvr or constntly decreasing bhvr thinking. ok. but stll i wnt to invstgte in arc lngth dmn to anlyse fgre out sme indctns on d,(n plus 2m) bèh. or it cld be stll other mdls also nt just 2. ok arc lngth categrztns wld help to get to anlyze i thnk.
started day with opening my dream living room view&sounds whilst starting studying: https://www.youtube.com/watch?v=oakA7RLvmWs&t=1029s :)
ok. study time starts.
ok today the verlet issues would be resolved.
--------
i saw very interesting text somewhere newly :). this is very very very interesting.. i newly saw :) very very very very very interesting :)
i try to understand exactly what it means. (as told before i would be very happy to apply to a secure home providing place in west everywhen in security institutions like places that hires software security jobs or such places). i couldn't exactly know counterpart that written that for which purpose (that text). but i wanted to write what i thought as i read. this is very intriguing/interesting/exciting to read about:). maybe no one written it for me and it were just coincidence. i dont any surely know:) (but if i have no such job application chjances, as told i build a windturbine product to create a winter home to me. but if i have any chances i would be always happy to apply to surely:)) (i dont know if that were coincidence or not surely:) but wanted to mention my idea for that:))
--------
some project related topic i wanted to note:
arc-lenght based parameterization has nice properties for induction like: (for coplanar situation)
need to add also B (noncoplanaer case to account) but lets first try to solve these stuff if path goes coplanar to check the possible configuration dynamics relations in between dr/dt_over_n to dr/dt_over_n+2 for to fix verlet integrator's error exponentiation behaviors... (if B existed it would also enter this relations on its own dimension i think below happens on T and N plane. and if B existed similar below relations i think would happen on B and T & B and N planes but need to check this also. this is just an idea not proved. would need to check.)
---> if n is even: (-1)_over_(n/2) * T/r_over_n
/
dT/ds_over_n
\
-----> if n is odd: (-1)_over_((n-1)/2) * N/r_over_n
------- (i had written with some errs in writing from study page to blog page, fixed them now :)
maybe i try to move the integrator arc-lenght based system completely also with removing t from all and trying to fist figure out t lacking possible configurations of paths of such system. i need to recheck lennard james potential equations to investigate aha from a i could convert to arclength again. ok i would change the integrator to utilize mixture of arc-length and time based version to have chances to do the taylor more correctly. yupp i think i wish to finalize this this night. yayyy a better version of verlet integrator is getting built. with mixture of arc-length and t based interpretations yupp.
its still unsure if this topic i thought could be devised. i might figure out this method is not as easy to devise during the study today. lets see. wishfully i wish it could be devised. i think it would be. so that we have a better verlet integrator afterwards in every molecular simulator. i never got content with verlet integrator and have lots of objections to using in my simulator. it felt all wrong if i used verlet integrator in my simulator. i wish this night I resolve this integrator algorithm :) i open source my silly integrator method and would write a blog post about it. it would be named gusili integrator if i could manage to build this night (my cousins loved to call me gusili long before. i think it sounded good to me, gusili or guhu afre my short names in friends/cousins. cousins like to call me gusili. previous istanubl friends liked to call guhu. i think i liked gusili integrator version more than guhu integrator version. ) its yet unsure if this method of gusili integrator could be built or not. i wishfully build this night.
------
Yorumlar
Yorum Gönder