Meeting Title: Eden __ Brainforge - Python script review with Bo Date: 2025-01-09 Meeting participants: Luke Daque, Nicolas Sucari, Bo Yoon


WEBVTT

1 00:00:14.460 00:00:15.350 Nicolas Sucari: Hello!

2 00:00:18.130 00:00:19.030 Luke Daque: And you go.

3 00:00:23.460 00:00:28.569 Nicolas Sucari: You tired already this time today, I think. Yes, right. It’s too late for you.

4 00:00:28.570 00:00:36.549 Luke Daque: Yeah, a bit, especially considering I was like migrating a lot of yeah queries, it was pretty

5 00:00:38.110 00:00:42.000 Luke Daque: yeah, pretty tough, really.

6 00:00:42.530 00:00:43.650 Nicolas Sucari: Really long day.

7 00:00:44.050 00:00:44.980 Luke Daque: Yeah.

8 00:00:48.410 00:00:53.310 Nicolas Sucari: The idea here is both to just share a quick walkthrough of the code.

9 00:00:53.530 00:00:56.369 Nicolas Sucari: He he shared the code. You have you seen it?

10 00:00:56.370 00:00:59.700 Luke Daque: Yeah, yeah. But I haven’t really like taken time.

11 00:01:00.370 00:01:02.079 Luke Daque: Understand it. But I I just.

12 00:01:02.080 00:01:02.610 Nicolas Sucari: Thank you.

13 00:01:03.020 00:01:04.440 Luke Daque: Took a quick look.

14 00:01:07.090 00:01:08.760 Nicolas Sucari: Ideally with that.

15 00:01:08.780 00:01:09.810 Luke Daque: Far.

16 00:01:09.810 00:01:10.400 Nicolas Sucari: Yeah.

17 00:01:11.350 00:01:12.430 Luke Daque: Yeah. Yeah. Go ahead.

18 00:01:13.940 00:01:16.410 Nicolas Sucari: Ideally with that, we’ll just try to

19 00:01:16.610 00:01:22.800 Nicolas Sucari: replicate. Maybe it’s just like asking Chatgpt to translate that code into

20 00:01:23.741 00:01:27.619 Nicolas Sucari: Dbt file, or something like a SQL. File right.

21 00:01:27.620 00:01:34.940 Luke Daque: Yeah, but it looks like it might be a little more complicated like the other.

22 00:01:39.520 00:01:41.970 Luke Daque: So Hello!

23 00:01:42.420 00:01:43.679 Luke Daque: Can you hear me now?

24 00:01:44.190 00:01:46.540 Nicolas Sucari: I can hear you now. Yeah, I think I lost you for a while.

25 00:01:46.540 00:01:50.919 Luke Daque: Yeah, yeah, sorry. I think it’s my Internet connection. But

26 00:01:51.350 00:01:56.499 Luke Daque: yeah, there’s 2 files there. The main by, and the function.

27 00:01:56.670 00:01:59.210 Luke Daque: The main one might be pretty

28 00:01:59.340 00:02:02.594 Luke Daque: easy to do in in Dbt, but the other one

29 00:02:03.130 00:02:07.270 Luke Daque: like it, has like a lot of probably.

30 00:02:07.270 00:02:07.820 Bo Yoon: Hey, guys.

31 00:02:08.580 00:02:10.040 Luke Daque: Oh, Hi Hi moon!

32 00:02:11.170 00:02:11.490 Nicolas Sucari: Table.

33 00:02:12.120 00:02:13.140 Bo Yoon: Good afternoon.

34 00:02:13.170 00:02:13.980 Bo Yoon: How’s it going.

35 00:02:14.800 00:02:16.569 Luke Daque: Yeah. Doing. Well, how are you?

36 00:02:17.290 00:02:21.540 Bo Yoon: Good, good do you guys get a chance to look at the code that I sent you.

37 00:02:23.200 00:02:24.460 Nicolas Sucari: We were just talking about that.

38 00:02:25.810 00:02:26.170 Bo Yoon: Okay.

39 00:02:26.170 00:02:27.450 Luke Daque: Took a look at it. Now.

40 00:02:29.480 00:02:37.990 Bo Yoon: Yeah, I don’t. I don’t think it’s really complicated. It’s just we just gotta design a way to build like a pipeline connected with the.

41 00:02:38.280 00:02:40.290 Bo Yoon: It’s a SQL. Server that we have.

42 00:02:41.300 00:02:44.209 Bo Yoon: so I don’t know what. What do you guys think.

43 00:02:47.680 00:02:57.489 Luke Daque: yeah. Another way. We can probably do this if we can’t convert this into sequel, like, like, 1st thing that we will want to try is to see if we can

44 00:02:57.910 00:03:07.240 Luke Daque: use sequel for this transformation. But if it’s a little bit too complicated for sequel, then maybe we’ll just leave it as a python script. And maybe

45 00:03:07.740 00:03:14.529 Luke Daque: yeah, maybe maybe we can even run the script in a cloud function, a Google cloud function

46 00:03:14.700 00:03:20.530 Luke Daque: and just schedule it like maybe once a day, like we like you mentioned yesterday, and then it would

47 00:03:20.850 00:03:23.630 Luke Daque: still output a table in bigquery

48 00:03:23.750 00:03:25.470 Luke Daque: that way. You don’t have to like

49 00:03:25.640 00:03:28.260 Luke Daque: download Csv files or stuff like that.

50 00:03:29.040 00:03:30.610 Bo Yoon: Hmm, yeah, that’s.

51 00:03:30.610 00:03:41.972 Nicolas Sucari: Yeah, the goal, the the goal here is not to do any manual work once this is done. So look if like, if that’s a possibility. And we are not. You know, we don’t have like

52 00:03:42.450 00:03:51.480 Nicolas Sucari: a way of translating this python script into Dbt, okay, I mean, we should look into that Google script way.

53 00:03:53.800 00:04:03.249 Bo Yoon: Yeah. So I I think the the data processing script that I sent you. That’s I think that’s doable in sequel. I don’t. I don’t see why not?

54 00:04:04.170 00:04:09.549 Bo Yoon: The the only problem for that would be that. We’ll have to do that by product.

55 00:04:10.310 00:04:11.230 Bo Yoon: So.

56 00:04:11.780 00:04:13.600 Luke Daque: That’s 1 of the problems.

57 00:04:16.430 00:04:17.100 Bo Yoon: Yeah,

58 00:04:18.903 00:04:22.939 Luke Daque: So like we’re expecting to have one table per product.

59 00:04:25.780 00:04:27.520 Bo Yoon: Yeah, that that’s what I’m thinking. But

60 00:04:28.060 00:04:35.110 Bo Yoon: there’s a good way we can do it. not sure.

61 00:04:36.970 00:04:45.579 Luke Daque: Yeah, okay, no, no worries. Well, I’ll take a stab at this. I have to. Take some time to see. I’ll look into

62 00:04:46.060 00:04:46.800 Luke Daque: the script.

63 00:04:46.800 00:04:47.150 Bo Yoon: Yeah.

64 00:04:47.150 00:04:51.490 Luke Daque: And see if we can convert this to SQL, and then, yeah, so.

65 00:04:51.490 00:04:51.890 Bo Yoon: And.

66 00:04:52.366 00:04:56.789 Luke Daque: Yeah, if I have any like, we have any further questions regarding like, how we can.

67 00:04:58.511 00:05:03.939 Luke Daque: Materialize this by product and stuff like that. Then yeah, we’ll go from there.

68 00:05:04.830 00:05:08.330 Bo Yoon: Yeah, sure. Yeah, just just let me know if you need any help.

69 00:05:08.630 00:05:09.365 Bo Yoon: But

70 00:05:10.600 00:05:21.350 Bo Yoon: we’re on the same page that the that the main python code that I send you. That’s that’s not doable in Dvt, right? That’s just. It’s just gotta be in python right?

71 00:05:24.070 00:05:24.430 Nicolas Sucari: Nor.

72 00:05:24.430 00:05:25.110 Luke Daque: Same thing, again.

73 00:05:25.110 00:05:33.690 Nicolas Sucari: I I think no, I think that one is something that we can do in Python. The the difficult one is the one of all of the functions, I think.

74 00:05:34.140 00:05:34.750 Nicolas Sucari: Alright. Look!

75 00:05:34.750 00:05:43.770 Bo Yoon: Yeah, I I’m not sure if there’s a way we can convert the functions in sequel, because I’m using all these libraries from python.

76 00:05:44.000 00:05:45.369 Bo Yoon: Never heard that.

77 00:05:45.370 00:05:45.909 Luke Daque: There is.

78 00:05:45.950 00:05:46.470 Bo Yoon: That’s.

79 00:05:46.470 00:05:55.049 Luke Daque: We’re doing like cater plots. And yeah, this is basically creating the the graph right? Like, the

80 00:05:55.050 00:05:56.770 Luke Daque: yeah, creating the graph.

81 00:05:57.797 00:06:06.570 Bo Yoon: Using an exponential fit to basically predict the future as well. That’s yeah. It’s kinda hmm.

82 00:06:09.220 00:06:20.550 Nicolas Sucari: Is this just like all of these python is for that by visualizations that you’re doing or like, have you considered using mixed panel, or anything else.

83 00:06:21.340 00:06:29.615 Bo Yoon: Mix panel. I I have no experience with mixed panel, but when I when I had a conversation with Robert, he told me that

84 00:06:31.330 00:06:34.930 Bo Yoon: it might be doable, but it’s it’ll be kind of hard to do it.

85 00:06:35.570 00:06:36.060 Nicolas Sucari: Okay.

86 00:06:36.060 00:06:37.160 Bo Yoon: And mix panel.

87 00:06:38.100 00:06:41.090 Bo Yoon: Yeah? And and the the other problem is that

88 00:06:41.460 00:06:48.730 Bo Yoon: this is not it for this app? I’m gonna add a lot of more more features in it. It’s just this is just the beginning of it.

89 00:06:49.350 00:06:55.220 Bo Yoon: So I I’m pretty sure that it’s gonna it’s gonna get a lot more complicated

90 00:06:55.810 00:06:57.439 Bo Yoon: than what we have right now.

91 00:06:58.920 00:07:02.359 Bo Yoon: So a custom app.

92 00:07:02.660 00:07:08.410 Bo Yoon: so building a custom app deploying it on Google Cloud, that I think that’s the best solution that we can

93 00:07:08.820 00:07:10.910 Bo Yoon: we can have at the moment.

94 00:07:13.550 00:07:15.915 Bo Yoon: unless you guys have another idea.

95 00:07:17.440 00:07:20.540 Nicolas Sucari: No, I think we. We just need to look deeper into

96 00:07:21.400 00:07:23.180 Nicolas Sucari: these files and see what we can do. There.

97 00:07:24.530 00:07:27.050 Bo Yoon: Yeah, sure, sure.

98 00:07:31.110 00:07:39.600 Bo Yoon: yeah. And and also for now I’m, I built other models as well, but I didn’t have time to convert that in

99 00:07:39.860 00:07:45.249 Bo Yoon: like in an in an app. Yet. So I’m presenting that in an excel format.

100 00:07:46.300 00:07:49.699 Bo Yoon: So I’m gonna share that with you as well, but

101 00:07:50.340 00:07:57.559 Bo Yoon: our end goal will be transforming that excel excel, file, format, file into the cloud as well.

102 00:08:00.090 00:08:00.780 Luke Daque: Okay.

103 00:08:01.770 00:08:08.933 Bo Yoon: Yeah, sure, yeah. Do you guys have any other questions for that?

104 00:08:11.400 00:08:15.110 Nicolas Sucari: I don’t think so. Not not me, Luke.

105 00:08:15.110 00:08:20.759 Luke Daque: Yeah, I I think not at the moment, but once I take a look at this more.

106 00:08:20.990 00:08:26.299 Luke Daque: Once I get more time to take a look at this, and maybe I’ll have some other questions. I’ll just probably

107 00:08:27.315 00:08:28.070 Luke Daque: contact.

108 00:08:28.070 00:08:31.659 Bo Yoon: Yeah. Just send me a slack message. Yeah, sounds great.

109 00:08:32.070 00:08:32.450 Nicolas Sucari: Perfect.

110 00:08:32.450 00:08:35.649 Bo Yoon: Yeah. Alright. Then, yeah, thank you so much for your time.

111 00:08:36.480 00:08:38.160 Nicolas Sucari: Thank you both as well. Thank you, Luke.

112 00:08:38.169 00:08:38.539 Bo Yoon: Yeah.

113 00:08:38.539 00:08:39.919 Nicolas Sucari: Talk later. Bye-bye.

114 00:08:40.929 00:08:41.609 Bo Yoon: See you guys.