How To Analyze CSV Files With Langchain And GPT the Easy Way

This content is generated from the audio voiceover so it may contain errors.

(00:00) very good to all and to all and welcome to a video more than the canal of iSocialWeb in this case go to land also a practical case of ia of how speak or communicate us with our data what go to do is to do an export of some data of csv that in this case go it to do with screaming frog so that it was directly related with cathedral And how can through the natural language the use of a model of the language do surgeries on these data costs to do this the first that have
(00:27) done is to do a crawleo selective with screaming Frog here the focused from the point of view of the architecture web what have wanted is to configure the track removing me images and everything or was simply the internal and external links have created like this the project and afterwards inside the analysis of skull have marked the option of link score so that it do me a calculation of the distribution of the page range internal of this web costs once that have done this that have launched it on the web of iSocialWeb what it would be necessary to do is to export this to a
(00:57) csv okay can Call it linked internal csv what want to okay export it voucher once that have this now already go us to the Script that have mounted to do this communication go to use the lanche lanchain is a framework that is in Python and in javascript that allows us interactuar with big models of language in this case Open there is voucher go the models of Open AI then here the only that do Is to install lanchains to install Open AI do some calls to the people that does all this with the different libraries
(01:32) any to call to the csv to pandas to show the information and use it for categorizarla and here what does the Script is very simple first asks us an archive csv that go to go up that already I have it prepared is this linked of internal voucher And once that we go up it processes it and asks us that we enter the question that want to do him in this case as it is a listing of urls that have a determined an internal, as we go him to ask Which are the 10 URL that are in the column direction in this case if you can give him
(02:06) the most defined information much better and that say us Which is go those that have more internal voucher more internal force then This does a call to the people of lanchain if it begins to do the process costs goes to begin to do the analysis Then already fix you that it does the surgery and then says Hears of the column of direction the type such And then here goes you ordering the urls as it has not finished to do the process the people goes to do more surgeries and goes to try #take data for the moment has #take of here
(02:37) already you see that it is doing a research of the ones of the data that more force have of the link score with the direction and then says you here the final response of the 10 URL that more than the link score taller that have are these of here #give to us ready format that have the 10 URL costs with this can go on doing questions of what want to okay here go putting new questions and go doing surgeries on the data this can extrapolate it to some mathematical documents to documents of analysis of finances to
(03:10) another type of crawleos to any type of data costs exports of tools of external like sixtrix, semrush or ahrefs what want to but that result you useful and see us in the next video