23OH-18-7-21.mp3 Harpreet: [00:00:06] Welcome to the comet Emelle Open Office Hours, powered by the artists of Data Science. It is Sunday, July 18th, 2012. I can't believe it is like over halfway through the year already mad. Feels like feels like stuff's just been getting started. Superexcited have all of you guys here. Hopefully you've had a chance to tune into the podcast or recent episode on Friday with the one and only James Altucher. That was a cool episode. I really enjoyed speaking with James. He's kind of a hero of mine. So it was it was beyond crazy to to get him Harpreet: [00:00:39] On to the show, though. The way I Harpreet: [00:00:40] Made that happen was interesting. Read James Olsher is all about ideas and being like the idea Harpreet: [00:00:44] Machine and write down Harpreet: [00:00:46] 10 ideas a day. So then I sent him an email just because he shouts out his email at the beginning of every podcast he does was like, alright, well I'll just send you an email. Um, and I said, James, here's ten things we could talk about. If you come on my podcast and Harpreet: [00:01:01] By the way, I'll be sending Harpreet: [00:01:03] You 10 ideas a day for ten days and Harpreet: [00:01:06] That, you know, you could Harpreet: [00:01:07] Block me if you want, but it's happening and like email. And he's like, all right, this sounds good. I'll come on your podcast. And it worked out. And I had James on the show. That was freaking awesome. So, yeah, I'd definitely check that episode out. Everybody listened in on Harpreet: [00:01:22] Linkedin, on YouTube, Harpreet: [00:01:24] On Twitch. You guys, if you have any questions, feel free to drop that right there into the chat. Be happy to take on any questions. There is also a link to get into the Xoom room right Harpreet: [00:01:36] There in the Harpreet: [00:01:39] Title or header of Harpreet: [00:01:40] The episode that is Harpreet: [00:01:42] Being streamed right now. So go ahead, click on that and come join us. Let's go ahead and get started off. You know how I like to do a kicking off with just like an open ended question just to get the conversation going. And obviously, you know, we'll be taking all your questions as well. But let's talk about so so chronobiology, I think, is an Harpreet: [00:01:59] Actual [00:02:00] actual Harpreet: [00:02:00] Thing, right, where people tend to work best during certain hours of the day or tend to be most productive or creative during certain hours of the day. I'm wondering for you guys that are joining us here, when is that for you? When do you tend to be the most kind of productive or creative and, you know, just feel like you're doing your best work? Let's let's start with that Cristoff here and see what he says. Cristoff: [00:02:25] Now, I think productive in terms of like work, like focus, work. Work is, uh, for me is Harpreet: [00:02:36] In the morning. Oh, it's Cristoff: [00:02:38] Really five a.m.. Harpreet: [00:02:40] Nine a.m.. I mean, I Cristoff: [00:02:41] Try to wake up at four to get started and at five Harpreet: [00:02:47] I start my Cristoff: [00:02:49] Really focused work. This is when I work on projects, uh, when I return. But I read about it when we've got really this context, uh, we get creative on the opposite time of the day. So it's like when you work focused in the morning, it means you're going to be more creative in the evening. And I guess I, I believe that. So I'd say deep work in the morning. Creative Day. Harpreet: [00:03:19] So that was this book that you talk about. The Chronobiology thing was that Daniel Pink's book? Cristoff: [00:03:23] When I know I've read it in in another book. Harpreet: [00:03:28] Ok, interesting. Yeah. Yeah. That's that's awesome that I tend to be the same way. What about you Austin. Austin: [00:03:34] I'm not quite Sivam but definitely also the morning. I'm not getting another that for yet. But no, I think for me it really stems from this like natural light, like just having light is very important for my productivity. I'm happy to be like I was in a basement apartment for four years. I mean, like a apartment with actual light, which is great. And also something about I think I do my best work when I have some sort of sense of external accountability. And [00:04:00] I think, like in the mornings feels like everyone's a little more active, a little more activated, like getting to work, doing their thing. Like it just it just gives me that little extra boost then kind of do my own thing. I think Christoph is right. I think I come up with my most associative stuff in the evening. That's like a less connected to a task less or less connected to my day to day. I just I think that's kind of an interesting point about the creativity coming later after you've expended all the energy to do the task. So I would really do that as well. Harpreet: [00:04:26] Yeah, yeah. Nice for me. It's also like I like waking up early in the morning, but then I use it early in the morning time to kind of do like my my own thing. That's the reading, writing, reflecting whatever side stuff that I that I've got to do. But I find that I start getting like really kind of more productive actually like around the mid morning, like right around like 10:00 a.m. to like one really solid, just like getting stuff done. Then I'll go outside for a walk and then I'll come back and, you know, handle a few things. Um, but that's kind of that's how it tends to work for me from either twenty two or ten to one. That's when I'm like like really in the zone and just getting stuff done. How about if anybody else. Shirley, let us know, so a partner or coach will let us know, by the way, if anybody has questions related to Harpreet: [00:05:10] Anything Data science, Harpreet: [00:05:11] Machine learning, breaking in Data science, working on the science problems, machine learning problems, whatever. We're here for those questions, as well as whatever other questions you got. We are here for you, my friends. But what Harpreet: [00:05:22] About Paratha or or Harpreet: [00:05:24] Ghoshal whereby you guys you guys kind of have a time where you feel most productive. Bharat: [00:05:31] So in my case, I'm still finding out what I was productive for me actually not being a fixed in certain circumstances. So I'm still on that process. Harpreet: [00:05:45] Yeah. You know, it helps just to get like a structure in place. I just really like structure kind of day. Right. For me, it's like early in the morning from whenever I wake up, sometimes it's it's four thirty, sometimes it's like [00:06:00] five. But from that early morning up until about like eight. Eight, eight thirty, that's when it's like really use that time to do my reading and whatever writing reflecting I had to do. Then I'll take a walk, then I'll come back in and I just every day is the same, like I have these chunks of time and just helps to I, I don't know about it, just like it just propels that, that productivity forward. If I just have that structure in place, I have my day broken out into chunks like that. So maybe try that man. Got it. Yeah. So, so you're asking me if I got any like. And Jim quick on the podcast, man, I didn't even try and he started Harpreet: [00:06:39] Ignoring my Harpreet: [00:06:40] Content after a while, like for the first like seven or eight posts, he was like liking it and sharing it and then he just stopped engaging with it. But gimcrack if you're listening, which probably aren't. Um, I got to get you on the show, man. Like, you have no idea how much your stuff has been beneficial and impacted me. Like, I love all this course material. I've taken all of this like online courses, like quick thinking, speed reading course. Harpreet: [00:07:07] There's a course on Harpreet: [00:07:08] Memory, bunch of different courses. Super interesting. Sobashima and my good friend Subash, good to see you here. Let's jump into a question here from things culture that has a question. Kozel, go for it. Are you still around? Yes, you are. Go for it. I'll just read your question here. Oh, no. Go for it. Go for it. You're unneeded. Bharat: [00:07:27] Yeah. So, hi. How are you going to get. So my question Harpreet: [00:07:32] Was like, oh, Bharat: [00:07:33] I have a limitation Harpreet: [00:07:36] Of a tool Bharat: [00:07:37] Like I can perform capacity and not at all like we can do it in Python using other models. What are my postures to perform that and basically and Microsoft Excel or Google Sheets. So what do you think? Like, oh, I have no limitations of using models. So which would be the best model to forecast in [00:08:00] my, Harpreet: [00:08:02] Uh, man, I wish I wish I was David Lingoes here man. Where's David when you need him. He's the Excel guru. I mean, I think in Excel he should be able to do RMI forecasting. I think Oremus kind of like that. That would be my go to when Data is seasonal and cyclical. Um, I would go to a rhema. I don't know whether that's available or not in Excel. Um, have you tried to see if it is. Bharat: [00:08:25] No, it is not available. Harpreet: [00:08:28] What do they have available in Excel. Bharat: [00:08:30] They just have experience more things than normal moving average. Harpreet: [00:08:35] Yeah. Let's see if I could find you, uh, some, some resources. This is how I would go about doing it. So this is going to be a live, uh, me trying to find information type of thing. So let's go ahead and do this. Right. So let's check this out. Um, so we're going to Harpreet: [00:08:50] Go, uh. Harpreet: [00:08:53] Microsoft Excel and let's do forecasting and seasonal, uh, then we can say formulas form. Let's see what we can find here. Harpreet: [00:09:07] Uh, linear and exponential smoothing. Harpreet: [00:09:09] That's the only ones available in Excel. Uh, yeah. Times, raves, forecasting Excel. Harpreet: [00:09:15] Uh, well, Harpreet: [00:09:16] Alex, maybe I got a good it's pretty recent to less than a year ago, so let's check this out. I'll probably link you to this thing right here. And, uh, that's probably well, look right here calculates statistics using formulas in Excel Windows. These two might be useful for you. David Langer, if you're listening, um, help us figure out how to do a time series in Excel. That would be very helpful. But I can link you to these to these might be helpful for you. Um, and. Yeah. Have you come across this article yet. A year. Oh you have. You have. All right. Well then, uh then this might be a little bit a little bit more helpful. 52 pages of it's an Excel though, so that's [00:10:00] why I couldn't be more helpful with that. I don't use Excel too much Bharat: [00:10:03] Like I do. I use a lot. But this is the one I got. Harpreet: [00:10:10] Yeah. If anybody listening on any one of the streams has Harpreet: [00:10:15] Any insight, Harpreet: [00:10:17] Be happy to, uh, Harpreet: [00:10:19] To, to hear that. Harpreet: [00:10:20] So if anybody, even here in the room has any insight on that, please let us know. Um but yeah. Sorry uh sorry. You don't have to have too much more for you there. Um yeah. If anybody has questions, go ahead and let let me uh let me know. I'm looking at all the streams. Don't see any questions coming in, but we're happy to take all of your questions. Everybody I LinkedIn I see there's a there's about a dozen of you guys watching on LinkedIn Tau is sitting in, uh, watching us on on LinkedIn. He wants to know what are we trying to do. What we're trying to do together is trying to do, um, essentially Time series forecasting in Excel, uh, with data that exhibits seasonality and cyclicality. Um, so that is the question that we are trying to, uh, to answer. So if you could join us in here, that would be great. Or if you can comment right there on. Uh LinkedIn That would be helpful as well. Daryn, you got a question. Go for it. Speaker6: [00:11:21] I prayed about you, my friend. I missed you. Harpreet: [00:11:24] I know it's been a while. How you been? Speaker6: [00:11:25] As good. I've been all preparing because of all this interview and assignments. I got so crazy tired and finally what one offered me. But I was lucky because this this one was just after first interview without technical assessment, because all technical assessment I failed, whether it's a Data science or Data analyst. Harpreet: [00:11:45] But this was Speaker6: [00:11:46] Just first interview. I think they just like me and they hired me. But I'll be the only Data analyst because the guy who got promoted will be product owner. So I'm just trying to give you a general idea of the what the position is. Basically, I'll be the only data analyst [00:12:00] which is managing the whole data and presenting its kind of API. Also, I really need your help, guys, when I go there. And but they probably taking some kind of crazy expert guys coming. Harpreet: [00:12:11] So you need help because they expect good SQL by your scores is not that strong. Well, the good thing is my sequel's not not too difficult to learn. I mean, you know, I did an entire school course. He understands dream job. So go ahead, check that course out. You are a Data and seem have students have got access to that school from the ground up. Definitely go through through that. That'll be beneficial. And for everybody listening, I'll actually be teaching a school class live next weekend for the Desk Virtual Conference. I keep forgetting about that thing. Guys, I need to start pushing that promotion that more next week. Go virtual. I will not only be emceeing the event, but also be teaching a school series. Sorry, guys, I got so much interesting stuff going on. I just forget everything that I'm involved in sometimes, but that's what I would say. So to get good at at school doesn't take too much effort. Go through the school from the ground up series that we've got on, on Data says dream job or anybody that's listening. That's not good at school but wants to get good at school for freeish, go to DCO Virtual. We teach an entire day about about school, so definitely check that out. Speaker6: [00:13:23] But any other suggestions besides like like when I go, how should I start, you know, the beginner discount at all because it's different from what I do right now. Harpreet: [00:13:32] Yeah. So any, any time you get a job or any job that you're starting, the number one book I recommend to everyone is the book called The First Ninety Days. That's a really, really good book about just how to kind of, uh, manage expectations, set set expectations, build a good relationship with your direct report and things like that. So definitely, Harpreet: [00:13:54] Definitely get that book. Harpreet: [00:13:55] It's a short read. I think it's it's definitely honorable. That's how I consumed it. And I think [00:14:00] it's like three or four hours on audible. So if you listen to it, I like double speak to get through it. It just, you know, one commute pretty much. So definitely check that out. Um, and in terms of just general like. You don't have to be like a superstar rock star here on day one, right, there's going to be time for you to go and get familiar with the company, get familiar with the Data, get familiar with the processes and the people. So I would just focus mostly first week or two, just getting to know who it is that you're working Harpreet: [00:14:30] With, getting to know Harpreet: [00:14:32] Whatever database administrators they have there or Data people they have there and just pick their brains on. Where does the Data live? And the company is a dictionary. Um, what what Data is, you know, kind of the most important for the company. What is updated most frequently, like a sense of the architecture is Harpreet: [00:14:51] There is it just one Harpreet: [00:14:53] Transactional like database and a bunch of tables and they have like a data warehouse. Is there a pipeline that goes from raw data to the data warehouse, or is that something that you're going to have to build like e-mail process? Do they have any detailed processes already in place? Are there any stored procedures? If so, can you see the documentation behind that? Because it would be helpful for you to understand what's happening during that Eataly pipeline. Speaker6: [00:15:16] Sorry, Harp interrupt you. Thank you very much. Could you please tell me briefly what is like pipeline? And it'll think Harpreet: [00:15:25] The Harpreet: [00:15:26] Eataly extract, transform and load. So pipeline means. Okay, here's Data in one place, right in one place here. You need to get it to another place. So you need to write code that will essentially extract the data from one place, do some stuff to it. The transformations and that stuff could be doing like aggregation or summarization or whatever it is that you need to do Harpreet: [00:15:49] To get it Harpreet: [00:15:50] Ready for reporting and then loading it to another either hopefully data warehouse Harpreet: [00:15:55] Or some other place. So that's Harpreet: [00:15:57] Etl. And the pipeline is just it just [00:16:00] code that makes it go Austin: [00:16:02] From end to end. Speaker6: [00:16:04] Do we do the analysis, do that or it's basically the engineer's job. Harpreet: [00:16:07] Uh, if you're the first Data person they've hired, it's probably on your plate as well. So when you're the first Data person hired and you kind of have to do a lot of different stuff and a lot of different hats. Um, but I mean, that's why I said get in touch with, like a database administrator. If they have one there or somebody that you can talk to, hopefully they'll get Data architect or something. But yeah, um, Speaker6: [00:16:31] I guess most of the things that built in already I just have to go and probably continue. Harpreet: [00:16:37] I got no clue what the company, uh, is like that it's going to be working on. So, uh, just talk to people and just talk to people and get a sense of what they need. If anything, one thing I would recommend in terms of just a tool to use when it comes to databases, my favorite tool for working with SQL database is called it's Azure Data Studio. And Data is nice because, um, the hardest part about SQL for me is, um. Imagining the transformation is in my head, like what's happening to the Data, AIs and writing the query, but with a notebook you can see iteratively, it's like, OK, if I do this and this happens and this happens. So Azur Data studio has like notebooks for SQL. So it's quite, quite helpful. A lot. Yeah. Um, I python notebooks but they, you can write SQL in them and you can even toggle between like, you know, python cells and single cells. It's really, really good. Speaker6: [00:17:36] Well, thank you very much. Harpreet: [00:17:37] Yeah, place equals Data sort of place equals destination. Yes, place to place, source to destination. Harpreet: [00:17:43] Uh, that's Harpreet: [00:17:44] The question coming in from the chat here in LinkedIn place equals datasource. Yes, there's datasource and there's a Data destination in these places, you know. Right. A pipeline to go from one place to another place. Harpreet: [00:17:57] But yeah. Harpreet: [00:17:58] Good good luck on the new road I think [00:18:00] in. Yeah. I know you've been working hard to, uh, to land this, so thank you very much. Looking forward to seeing what happens. Can you still register for the disco virtual conference. Absolutely. Harpreet: [00:18:09] You can definitely go Harpreet: [00:18:11] And register for DSTO virtual conference. I think it's just that you have to get the link for you. But if you just go DSTO virtual on Google and say July 20, Harpreet: [00:18:20] 21, you'll get the link that Harpreet: [00:18:22] Pops up and it's going to be such a good event. Uh, there's I mean, obviously I'm emceeing the entire thing, so, of course, it's going to be awesome. But there's also a panel discussion hosted by my good friend Kenji. And Kenji is hosting a panel about the podcasters of Data science, which Harpreet: [00:18:38] Conspicuously is missing me. Harpreet: [00:18:41] But that's because I'm emceeing the entire thing about his things and be cool. He's got a bunch of awesome podcasters that'll be on his panel. But then the courses that are taught the next day, I think the courses are taught on Sunday. So I have to figure out what's happening to officers next week, but we'll sort that out. But there's courses with John Crohn, who's doing a course that is Tensorflow versus PI Torch Team Patchworks. I like my torch. Uh, uh, Andrew Johns, Data Science Infinity. He is doing a, uh, entire workshop. I think he's doing like N10 Machine Learning. Joe Reise is doing a Data engineering session. And there's so much good stuff happening, uh, for Desco Virtual. I highly recommend it. Um, go and check that out. Um, and then the networking events, those are a lot of fun. I always like I always like to do the networking events. I just jump into those and, uh, see who I can run into. A lot of fun. Ashar, it's going on. Good to see here again. Um, anybody else ask questions? Go ahead and let us know into the Harpreet: [00:19:45] Chat Harpreet: [00:19:45] Wherever you are. There's some questions coming in regarding the Excel topic that, um, I think part of that asking about was Harp. The coach was asking about, um, on LinkedIn. Uh, yeah. Kozel on LinkedIn. Check out the comments. [00:20:00] Rodney has got some links and therefore you, um, create a forecast and Excel. He's got a link there from Microsoft documentation. So go ahead and check that out. Uh, Cenovus was asking, what's the book title of the book I was talking about with respect to starting your first job and kind of how to get on the right foot. That's called the first ninety days. The first ninety days. Great book. Um, yes. Go ahead. Let's take questions, guys. Um, I'm definitely open to taking questions. Shout out to everybody in the room. If you guys have questions, let me know. Um, everybody watch on LinkedIn. Those guys watching on LinkedIn. There's there's a lot of you on LinkedIn. So come and join us in the room. I like making this interactive as much as possible. Um, but yeah, of course we've got a lot of great, Harpreet: [00:20:45] Uh, tips for you on the LinkedIn comment section for your particular Harpreet: [00:20:50] Problem of doing forecasting in Excel. Harpreet: [00:20:53] I say that's quite interesting. Harpreet: [00:20:55] Get up. You have there. What is what is going on. Austin: [00:20:59] I'm not feeling too well. I'm trying to stay as warm as possible. You don't want to be on me? Harpreet: [00:21:06] Oh, man, I will hopefully feel what we feel better now than, uh, any questions, any questions or anything that could help you with. Because I know you always got some good ones. Austin: [00:21:16] I've heard you talking about the life forecasting in exile. Yes, I do. You don't want to die. Delta is something I came across this week I never knew existed. There's something called profit Harpreet: [00:21:29] O Facebook's profit package. Yeah. Yeah. Um, yeah, that's a good one as well. So I think Cowsills Constraint was here to only work out of Excel. But can't you integrate Python into Excel nowadays? Um, that's what I thought. That was the thing that is possible. Maybe that's Python into power by our Facebook profit is definitely good package for doing Time series. They came up with another one as well. Uh, recently. I can't remember what it's called. Um. But yeah, [00:22:00] so how's everybody doing? How's everybody's anybody got plans for this upcoming week? Anything interesting you're working on anything exciting or fun? I've I've had a crazy week, man. Just a Harpreet: [00:22:14] Ton Harpreet: [00:22:14] Of ton of interviews I've been on. Like, I think there's four interviews this week, bunch of take home assignments. That's just Harpreet: [00:22:20] Been my Harpreet: [00:22:21] Evenings have been all. Like, just tied up, so I'm excited to kind of get get past this little little bump, I got a couple Bharat: [00:22:29] More for the 20. Harpreet: [00:22:33] So definitely looking forward to to chillin out a little bit after this crazy, crazy week. How about you guys has has a race week looking coming up. Austin: [00:22:44] And can I go, Harpreet: [00:22:45] Yes, please. I think Asia froze up on us here. Yes, I should it frees up once you are back, I sure we're happy to take your question shot at everybody else in the room. Roett, I know you're sitting here, um, had your microphone muted. So if you have a question, please do let us know. Happy to take any and all questions. You guys are going to make it interactive session. There can be some questions. Yes. Yes. Go ahead and say there can be no office hours without any interaction. Otherwise, I just Cristoff: [00:23:20] I just don't know which one to Harpreet: [00:23:22] Ask. Well, let's go. Cristoff: [00:23:25] What's like the worst career advice Harpreet: [00:23:28] That you keep seeing Cristoff: [00:23:29] Or hearing or something like Harpreet: [00:23:31] Somebody is trying Cristoff: [00:23:33] To help or he's thinking he's helping. But in fact, there's no idea. Harpreet: [00:23:40] There's a lot of that out there, a lot of a lot of bad career advice. Um, I don't know, man, what's the worst bit of career? But I'd have to think on that because there's so much so much bad advice to what some bad advice that you think people are peddling out there when it comes to career. Bharat: [00:23:57] Putting me on the spot, I guess, I [00:24:00] don't know when it comes to advice given I never really had any bad advice, but I think he has to follow the advice, not really thinking about it, because everybody is giving advice based on their own standpoint. And I think the biggest danger isn't so much the prize itself. That's more how you follow what you take it into account or just kind of follow without thinking about it. Plenty of advice has been given all the time. Someone followed some, but you have to kind of think about it's always nice to get advice, but like I said, you always have to keep in mind that whoever is giving it has their own situation, background and their own reasons for doing that type of advice. Harpreet: [00:24:53] Yeah, absolutely. Ask them what some bad career advice you've you've seen or heard of recently. Austin: [00:25:01] That's a really interesting question, and it is very informed by where you're coming from culturally, because for me, I got a lot of advice about from when I was a kid, like my dad sort of, or my parents didn't really care much about, like, accumulating and passing on wealth. So I got, like, a lot of the perspective, just sort of like but I think I think maybe one is like turn this sort of blind thing of like turn what you're passionate about into a career. Like, I think that can work for some people. But like so for example, like I was I was in grad school, I wrote a book of poems, I was an MFA from poetry. And I saw the reality of turning that passion for creativity and associative writing and language into a career specifically like teaching poetry for a living or being a university professor and selling books and all these kinds of things. And that started sucking the joy out of it that I have. So I think the point that I was making about, you know, [00:26:00] contextualizing that advice to your own circumstances, like for me, I can find some of the things I enjoy in my work that I do now, and that's communication, community, all these things. But I'm not like if I had to write scripts like screenplays for a living or these things I tossed around in my head, I think I would lose the joy of it. So it's like it's it's sort of balancing, like finding things you're good at and can get really much better and turning to a career versus like following your passion, I think there's more nuance and that sort of career advice that is often given credence. That's what I would say. Harpreet: [00:26:33] Yeah. Yeah. And not given some time to think. I need to take into consideration some of what I'm doing. And I feel like they're the worst career advice I hear if blanket statements that people just assume will apply to everyone, I think any advice given needs to be definitely within within context. So that's why any time somebody comes to me for advice, like, I don't just tell them to do something, like I always come back at least three or four questions. I let me dig a little bit deeper, let me understand a little bit more, and then give you something that's going to be suited or tailored kind of for you or kind of more relevant to you. It's when you start trying to use advice as like a prescription that it starts to not work. Right. I don't know if that made sense, but Austin: [00:27:18] In other words, that's really smart. That's really smart. I think that that's that's sort of it's born of like the sort of ways in which we communicate on these platforms now that almost animals are more prone to or more amenable to that sort of blanket statement and lacking sort of follow context for the questions you need to ask yourself to begin to like, is that right for me or at what context is that right for me? So I think that's a great yeah. Harpreet: [00:27:43] And a lot of I would say there's a lot of bad advice starts with why don't you just or just do or just this or just that like like as if it was that easy or you want to be. It is why don't you just go and do these few things and even even though I'm like, oh God, just do a project. I mean there's a lot [00:28:00] more to it than that. So that's why I always try to try to say, yes, do a Harpreet: [00:28:05] Project, but make sure you do Harpreet: [00:28:07] It right. But yeah, a lot of the worst advice I think started with words. Just what about what some really horrible advice that that you've gotten. Austin: [00:28:17] Just received a lot of advice that day, were all good, I haven't received a lot of advice, especially with transitioning from individuals, but from definitely from online. It's what you said, the blanket statements you can do it in for months. Just concentrate two weeks. You'll be done. That's the worst. How do you get a timeline of things? Harpreet: [00:28:38] Yeah. And I mean, I think the reason you don't get bad advice is because you give really good context like this. My situation this way, I'm working on it. So I'm trying to do. Harpreet: [00:28:46] Do you have any advice from me? How do I move a step Harpreet: [00:28:48] Forward or what have you? Um, so I think that's probably why you tend to get good advice herself. Was that was that helpful? What have you what have you seen out there that's been really, really just not you know, I Cristoff: [00:29:03] Just reminded me of Harpreet: [00:29:05] I it was like two months ago Cristoff: [00:29:07] Maybe when Andrew Johns had this post on LinkedIn that started Harpreet: [00:29:13] That I Cristoff: [00:29:14] Keep hearing this. You're not a real data scientist unless you something you you can do. Well, you do that. You do that. And I think that's that's like really negative content that we find. And and it's it would be very useful to Harpreet: [00:29:39] Just block Cristoff: [00:29:40] This kind of content because it brings only negativity. And this is very biased by people who write it because they tend to give more importance to things that they already know. I mean, because they like to compare [00:30:00] themselves with others and they like to think about themselves like they better than others, because they know that Harpreet: [00:30:08] It's it's more Cristoff: [00:30:09] Important because because they know that it's like this kind of B.S. that makes Harpreet: [00:30:16] Something that it's supposed Cristoff: [00:30:18] To be an advice, but it only brings negativity and like pressure on people who are starting Harpreet: [00:30:24] And they think Cristoff: [00:30:26] They are too slow or not smart enough or they're missing something, not job ready. Harpreet: [00:30:32] Something I do 100 percent agree. I hate those type of posts. I remember, dude, like years ago, many years ago, maybe like two, maybe three years ago. And this post sticks out in my mind. And it was somebody that was like I respected a lot at the time. Like this is like an O.G. in the field. And they posted this asinine comment, which was you're not a real data scientist. If you haven't been earning at least six figures for three years, by the time you've turned thirty, I'm like, what the fuck? Are you serious, man? Like, that is the stupidest piece. Like, what are you talking about? I don't know if they're being facetious or if they were just that's actually how they felt knowing this person's track record and how they have been Harpreet: [00:31:11] Posting this, actually how this Harpreet: [00:31:12] Person probably felt Harpreet: [00:31:14] And that I was like, did you like you're a person that people Harpreet: [00:31:16] Look up to a lot like you can't be saying shit like this. Let's go to let's go to Austin then. Austin: [00:31:24] Yeah. I just want to say one last thing on this real quick. I think I said something about the platforms. I think there's something about the platforms on which this advice is a lot of time shared where like things like that, that stir up controversy or stir up emotion, even if they're negative, are the things that get people more attention. So it's almost like there sometimes you have to tease out, like, who's really trying to help you versus who is trying to get your attention. And some people are going to doing both. And the best people I've found in the people I follow like Harp. You're an example of this. I know Marc, who sometimes joins us as an example of this. They're teaching us something through their advice. They're not just pitting people against each other. And it's so easy to get attention, [00:32:00] but pitting people against each other, saying things that are controversial. And it's like it's just like a bullshit detector. It's like you got to learn how to sniff that out and it gets easier over time to do that. But you have to really be aware of that because you can get sucked into that cycle of negativity so quickly and feel bad about yourself. You know, Harpreet: [00:32:16] The coach will go for it. Bharat: [00:32:19] Ah, so I like I have a question Harpreet: [00:32:22] That what Bharat: [00:32:24] Kind of questions I can ask, to be honest, Harpreet: [00:32:27] Everything, man. Anything there's like I mean, it's like anything Bharat: [00:32:32] I just want to know. Like we have any Excel guru over here, like we can even help me in some kind of leaning Harpreet: [00:32:39] Because, like, I Bharat: [00:32:41] Have a good hand in Python, but not in Excel. And I'm going through interviews and working on the assignments to submit them. So I'm facing hard time to do those aggregations. And I'm trying like since to day two days and two nights, I'm trying to find new things and I'm not able to find it. So, oh, if you allow me, can I share my screen to show what kind of problem facing. Harpreet: [00:33:06] Yeah, definitely. I will say, though, there's some great comments here on the LinkedIn comment section, particularly around the question you're asking. There's the tour's great at Excel. He just doing it right here. Also, like if you haven't already followed David Linga stuff, he has a bunch of free course content out there that's like it's all centered around Excel, but it's how to do like SQL, like operations in Excel. So it's quite useful in that regard. But let's do this, though. First, let me let me quickly go to my tool and then we'll jump to to you, because I feel like this is going to be a Harpreet: [00:33:41] Really Harpreet: [00:33:42] Involved thing. And I know my tools had his hand up here for a for a few minutes. So let's do his question real quick. I think he might have a comment on what we're talking about before and then we'll get right into into, uh, into yours. So. Bharat: [00:33:55] Mittagong. Yeah. Austin: [00:33:57] Hi there, sir. I actually had a question. [00:34:00] I don't know if you want to finish out this question before. Harpreet: [00:34:03] Uh, no, we'll take your question because that this this I feel like it will be a long and involved thing. They'll stuff. So if. Austin: [00:34:10] Yeah, so my question was, so I'm actually doing an interview for a senior role at a consulting company. Now, I've never worked in industry before. Harpreet: [00:34:19] I've been a Austin: [00:34:19] Data scientist in academia. So working at industry, there's one thing I have to do is deal with clients. So I wanted to know as a data scientist working in industry, how do you actually interact with clients Harpreet: [00:34:34] And what kind of things, you know, you would want to Austin: [00:34:37] Interact for and with clients? I guess the date is for getting Data or, you know, asking questions or what kind of what kind of things you look for. Harpreet: [00:34:48] The clients are typically hiring you because they need your help to solve a problem. So when it comes to interacting with the clients, you just need to figure out what their problem is, make sure you really understand their problem and not like half acidly think that you understand the problem and assume stuff. So just questions and questions, no questions to them as long as it's helping you Harpreet: [00:35:09] Frame what what Harpreet: [00:35:11] It is that you need to do. So like if if you're working at a consulting company, um, I mean, ask as many questions as you possibly can, get as much context around what it is they're trying to do. Don't be afraid to ask questions and don't be afraid to quickly kind of iterate on little solutions and not Iranian military solutions. Show them get get that feedback as much as possible because you don't want to spend like four weeks on something only to come back to the client and say, oh, my God, this dislike completely misses the mark. This isn't what I was asking for. It's better to have those micro touch points, whether it's, Harpreet: [00:35:52] You know, once every Harpreet: [00:35:54] Week or Harpreet: [00:35:55] Or Harpreet: [00:35:55] Twice a week, even just to make sure you're on on path. Right. Especially in consulting, [00:36:00] because your billable hours is definitely a thing. But you also may be like as efficient as possible with with your time. So asking good questions is going to be a superpower for sure. Austin: [00:36:13] Right. And I actually have a final one for that entry. So any any tips on the. Yeah. On the client side, if I get asked any questions on that or so. Harpreet: [00:36:28] Um, so Harpreet: [00:36:30] I mean, I don't know how much time you Harpreet: [00:36:32] Have, but there's, there's a book that I enjoy, um that I kind of find myself flipping to every now and then you might be able to find a PDF of it. It's called Case in Point and it's a book is written for a case study style interviews. It's not specific to Data science, but it's just how to kind of solve that solve and think about problems. Harpreet: [00:36:54] Um, so Harpreet: [00:36:56] That's a good reference to got a bunch of exercises. They got good frameworks for how to think about solving a problem with them, you know, consulting kind of context. So definitely check that out. Uh, what's, uh, toward the only advice when it comes to consulting for me? Oh no. Talk to me. Bharat: [00:37:16] I mean, I've been doing consulting. For years on any job working in this job, as far as I can tell, and what I think is that your goal in this situation is really to truly understand what the client wants. In most instances, clients do not know what they want. It's your job to ask the right questions, guide them, clarify so that they understand what the solution should be. And at the end of the day, if you get a client to actually be the solution that you cannot deliver, that's your goal. Harpreet: [00:37:57] Yeah, yeah, absolutely. When is your interview [00:38:00] coming up until Austin: [00:38:01] It's on Tuesday, Harpreet: [00:38:02] Ok. Yeah. So hopefully this is good advice. I don't know how much time I'd like, you know, if you've got time to check out case in point before Tuesday. But, um, like the thing is, you just got to be able to solve problems. Like, I just be able to ask questions, provide solutions, get feedback, incorporate the feedback into your solution plan and increment forward like just as much as he can increment. Forget the feedback. That's the biggest thing I think is just that feedback loop. Bharat: [00:38:32] Yeah. Yeah, yeah. Thanks. Thanks for all the advice. Yeah. Harpreet: [00:38:36] Hopefully that is a hope. That's helpful. Good luck on the interview. Looking forward to hearing some good stuff from it. Um, thanks. Yeah. So Kozel, Citigroup, we will get to your question. I just want to, uh, knock back a couple other questions here. Um, there's a question coming here from, uh, Mohammed. But what is a good strategy to learn machine learning or deep learning when there is a ton of resources out there and it is just overwhelming? I think a good strategy is just to pick one of those resources and just kind of go through it and then to that resource help you. Do you feel Harpreet: [00:39:11] More Harpreet: [00:39:13] Illuminated by the time you've gone through that resource where there are parts of that resource that you still have questions on? If so, then go address those specific questions. My whole thing is just a top down type of approach and, you know, sometimes jump in mixing Harpreet: [00:39:28] In with the bottom up approach. Harpreet: [00:39:29] So top down, you just quickly get an idea of the workflow of how you go from data to decisions. Right. Just work through something real quick at a high level. Right. Whether that's like a mini Harpreet: [00:39:40] Project or Harpreet: [00:39:41] A resource or a course or something, just go through it on one path and then go back through it and address the parts that you're still having questions on are still having confusion with. I think that's probably the best way to do it. Um. Yeah. Like, don't pick up every resource, just focus on on one. If [00:40:00] there's one resource I would recommend, um, I think the gold standard book I think for introduction to kind of machine learning is uh or really on Garance book. Harpreet: [00:40:13] I think that's I Harpreet: [00:40:14] Say his name and that's, um, introduction to Machine Learning with Python and Tensorflow I believe. Cristoff: [00:40:22] And it's called Hands on Machine Learning. Let's take it. Harpreet: [00:40:26] And that's the one. Yes. Hands on machine learning with secular and tensorflow. That's the good one. That's the best one. Yeah, that's a really good one. That's the book that I used that like when I was first kind of trying to both learn Python and machine learning simultaneously. Um, that was very, very helpful. Harpreet: [00:40:44] It's a good book. Definitely check that out. Um, but yeah, Harpreet: [00:40:47] It's, uh, most of learning is deciding what not to Harpreet: [00:40:51] Focus on. So just. Harpreet: [00:40:54] Pick one, ignore the rest. Work your way through that, and that's it, just don't focus on the tons of resources, just focus on one good resource and just go through it. Everything you don't understand. Mark it down, note it down. Like, OK, this is something I don't get. I have to come back and revisit it. Um, it's kind of work through it as much as possible. The resource I suggested for my tools and was called case in point. That's correct. Case in point. Um that's a good, good read. I like thumbing through that one. Um, all right. So let's go for it. Uh, Koshary still here I think. Bharat: [00:41:28] Yes I am. Harpreet: [00:41:30] Cool. Let's go further, see what your question is if you want to pull up your Excel thing. Bharat: [00:41:34] All right. So basically I have an Data which is of half and half an hour away. So you can see this at twelve a.m.. I have protested this much. I have type one issue to take this much. I have five to ensure that it does not. And this is the half hour interval. All [00:42:00] right. So what I want as an output is basically this. I want our interview with. So basically, I want to aggregate that those things this, too, right? Harpreet: [00:42:13] So this would Bharat: [00:42:14] Be added up to would be added up. They should be added up. So it will go into the zero hour of the day. I have done that in Python Harpreet: [00:42:24] And I have to look back into Bharat: [00:42:25] The Excel, but I'm not sure that how I can do the same, except I have tried to grow by use, go by and many more Harpreet: [00:42:36] Things like lots of different Bharat: [00:42:38] Things, lots of different approaches. But I believe there is some advanced Excel formulas, which I'm not aware of at all. They might help if anyone can help on this piece. Harpreet: [00:42:48] I will defer this one to to tour. I mean, I wouldn't know how to do that, except it's easy enough to do in Python. But how would you do that in Excel for any advice? Dave Langer, where you we need you. Bharat: [00:43:00] I think they would definitely know. Probably a better solution. I have. But to me, this is really about the far internal Data the column that you have. You would need to somehow break that down into two columns, probably by hour. And that usual round formula to calculate any half hour is up to the whole network. And this way you can then generate technically a new Data similar to what you've done in Python. Yes, I have used the same formula as well. I have Harpreet: [00:43:33] Like Bharat: [00:43:34] Taken out the Data and are differently and I have used it. But I want to aggregate this two Harpreet: [00:43:42] Things and I'm not Bharat: [00:43:43] Able to do that. I am able to do it. By which? By the table. But I'm not able to do a like unpiloted back and take it into the proper format because like when I'm going to present this, I have to show a proper procedure and [00:44:00] then tell you like how I have done because I can I have done it, that I would say like twists and turns, but I don't have any proper process to do this or any proper reform. That's. Let me ask a question just quickly. Is this something that you're going to be using as an ongoing basis of like time? OK, so I'm working on it and w oh, so I got this assignment and I have to prepare or analyze the data, but I did and I have to realize that and I have to present it in an interview. So I'm concerned that if I'm going to use any other other tools or skills or any other things, that it would be accepted or not. I'm not sure about it. I. That's a good question, but personally, I don't think the end result is what counts. I get there is a different story and how you approach it, the way I look at it is that if this is something that you would be then you would be doing on a regular basis, what you're looking here to do is to actually automate the process. And you're probably better off using Bitly.com/adsoh you have to actually generate Data and that you use the new data to actually generate an excel, which you need to create for them, forecasting, et cetera. Exactly. Yeah. I like. Yeah. OK, go ahead. Sorry. So once you have the new table basically you will just how old Data then you run it through the pipeline, then you get the new data set. And based on that new data set, this is what you. Then the question is going to be using Excel, which you are working to, then you definitely want to build a structure in itself that allows you to use, as I say, Harpreet: [00:45:57] How to Bharat: [00:45:58] That you don't have to be [00:46:00] manually going on constantly on a regular basis to update formulas, etc.. OK, the exact tool, Harpreet: [00:46:10] If I could jump in real quick, I would just stop the sharing here Harpreet: [00:46:13] And I'm just pull Harpreet: [00:46:15] Up something and see if I can do that here. Uh, how am I going to do that? Share screen. Give me one second here. So here's what I've come across real quick. Just a quick Google search. So I think what you're trying to do is called a power query, apparently. So this is Excel. Um, this other documentation. You can group same values in one or more columns into a single grouped row. And this is a kind of walk you through how to do that. So maybe the concept you want to look into is a power query in Excel and had a group of stuff so I can send you these and I think that might be helpful. Um, so this is kind of like the search term I looked for to get there. Uh, and so power query group aggregation using amp functions, uh, things like that. So, uh, PowerCore regrouping and summarizing Data. So uh, this might be what we're looking for. I'll go ahead and I'll post these links here. Um. But yeah, I think this will be helpful. So maybe look up that kind of language, that concept of power recovery. Bharat: [00:47:21] So thank you so much. Yeah. Thank you so much, guys. Harpreet: [00:47:26] Yeah, so hopefully this helps let us know how it turns out. Give us a follow up next week. Um, you know, drop these links right here. Power query group. That's what this thing is called. Um, yeah. Check that out. Few resources here. Bharat: [00:47:42] So much Harpreet: [00:47:43] Cool. Yeah, no problem. Any other questions for anybody. Let me know. Please do let me know. Got a few more minutes here. So I'm happy to. Any questions I'm looking on. Uh LinkedIn don't see any questions on LinkedIn or YouTube or twitch. Um, we [00:48:00] can start to wrap things down. I shall go for it. Austin: [00:48:04] So I've switched off my Harpreet: [00:48:08] Mind Austin: [00:48:09] A little bit. So this was making that stuff up. Yeah, no problem. So like I said, I think about yesterday or just yesterday. Friday, yes. I studied a new role, but I'm working with financial data and floating half the time, which is quite OK. I'll catch up very soon. But another thing I'm coming to realize, I haven't worked with this on top of what I had asked about risk, risk and financial risk management is customer analytics. And by this I mean exactly what I had asked. Again, behavioral models, if you don't buy anything, let me ask the question again. Do you have any tips and behavioral models? Because, I mean, it's dependent on the customer you're working on. Let me try to explain. It's any behavioral model depends on what you're trying to achieve and the Data you have the different variables you have. But I have a feeling that I also need to understand how customer analytics, what's the financial bit. So on top of the book you recommended yesterday. Is there anyone else I know you said you're working with a lot of the financial year in the financial field. Any books you can point me to in the financial analytics Bharat: [00:49:27] I could make? Unfortunately, I'm not a big book reader, so I don't have any suggestions on it. Sorry. Harpreet: [00:49:38] Yeah, I think that book that I would recommend that they might potentially be overkill, because it doesn't really do that customers like, Harpreet: [00:49:47] You know, behavior models, Harpreet: [00:49:48] Anything like that, that's just strictly like a risk and derivative pricing and stuff like that. Um, Rodney is here saying Lee Anthony. Oh, wait, no, that's something else. He's just responding [00:50:00] to to a comment here from from Lee Anthony about that. Cautious question. So, uh, I say at least look this up, let's look this up here, let's let's do a quick search. Austin: [00:50:15] Um, I've done the search of just we have been looking at the teetotalers. I just feel like they don't go deep enough. Harpreet: [00:50:20] Yeah, well, let me just just keep it surface level. Austin: [00:50:24] Just wait it out and the just. Harpreet: [00:50:27] Yeah. Or just ask questions for the people who are experts in your company about this type of stuff. Um, so take take that top down approach and then I go from the bottom up on points where you need to. Right. So is it useful for you to get kind of deep into the weeds right now Harpreet: [00:50:44] To kind of get the job done Harpreet: [00:50:45] Or get a result? If it's not, then can I keep that high level for now just to kind of produce a result and dig deeper when you know when it's time to? But I mean, obviously, it sounds like you're having issues, finding actual resources to get Austin: [00:51:00] That to get to take the approach of landing in the office. They think I'm the dumbest one in the group for now. Harpreet: [00:51:05] For now. Austin: [00:51:07] But it's definitely going to get better. Harpreet: [00:51:11] That's always that's always a good Harpreet: [00:51:13] Feeling for me. At least I Harpreet: [00:51:14] Like being the dumbest person in the room. That means that this is a place where I can grow and learn and get better. So depends on how you want to frame it. You never want to be the smartest person in the room, that's for sure. Or you never want to think that the smartest person in the room. But, uh, don't worry about that. Like you you got an entire career ahead of you to learn and and do stuff. So I feel bad about that. Austin: [00:51:37] All right. Thank you so Harpreet: [00:51:39] Much. Awesome. Harpreet: [00:51:41] But yeah, I mean, shout out on, um, if you're in slack, um, you know, the hours a day science like community tag that community there with with the question of resources you're looking for and also be a reminder to me to Harpreet: [00:51:55] Help and dig Harpreet: [00:51:56] Up whatever I can for you as well. Um, and again, [00:52:00] I'm interested in seeing how this turns out for you. Harpreet: [00:52:03] Thank you. Austin: [00:52:03] To survive the first few months. Harpreet: [00:52:05] But thank you. Harpreet: [00:52:06] Yeah, no problem. Um, I see no the questions here anywhere actually as we can begin to. Oh Christof, you had a question. I saw you had your hand raised, uh, Harpreet: [00:52:17] Just a couple of minutes ago. Cristoff: [00:52:18] Just a quick one. Uh, have you heard of this book? Harpreet: [00:52:22] Deep Thinking. I've never heard of that one by Gary. Oh, that's the chess master. Gary. Cristoff: [00:52:28] Yeah, yeah. I got it today. So, you know, like every book. Harpreet: [00:52:32] So, you know, it's interesting. Harpreet: [00:52:37] No, I know Garry Kasparov. I know he wrote a book and, uh, deep thinking. All right. Let me know what you think about that man. I would like Harpreet: [00:52:44] To Cristoff: [00:52:45] I like this subtitle. It's called Where Artificial Intelligence Harpreet: [00:52:49] And Human Creativity Pilkey Harpreet: [00:52:51] Aufiero. All right. Well, I think I'll buy that right after this call. I'll definitely, definitely pick that up or at least download it on audible. Definitely. Good recommendation. I'll check that one out. I guys will go ahead. We'll wrap it up again. Thanks for joining us. Thanks for sticking with us. Till the end. Everybody on LinkedIn. Hope you guys eventually join us at some point as good. Having everybody here remember to, uh, tune into the podcast, release an episode with Harpreet: [00:53:18] James freaking out. Harpreet: [00:53:19] Altucher So let me know what you guys think about that. That was an episode that I was super, super hyped for. So definitely check that episode out next. We got an episode being released with John Fiola, who is the author of or coauthor of The Self-made Billionaire Effec. Another episode being released with Jonathan Tesser, who's kind of big on LinkedIn LinkedIn famous, and then one with Lillian Pearson as well. So finally, a data scientist back in the mix. You guys take care of the rest of the weekend. Remember, you've got one life on this planet. Why not try to do some big COVID?