Regression Intro - Practical Machine Learning Tutorial with Python p.2
Вставка
- Опубліковано 30 вер 2024
- To begin, what is regression in terms of us using it with machine learning? The goal is to take continuous data, find the equation that best fits the data, and be able forecast out a specific value. With simple linear regression, you are just simply doing this by creating a best fit line.
From here, we can use the equation of that line to forecast out into the future, where the 'date' is the x-axis, what the price will be.
A popular use with regression is to predict stock prices. This is done because we are considering the fluidity of price over time, and attempting to forecast the next fluid price in the future using a continuous dataset.
Regression is a form of supervised machine learning, which is where the scientist teaches the machine by showing it features and then showing it was the correct answer is, over and over, to teach the machine. Once the machine is taught, the scientist will usually "test" the machine on some unseen data, where the scientist still knows what the correct answer is, but the machine doesn't. The machine's answers are compared to the known answers, and the machine's accuracy can be measured. If the accuracy is high enough, the scientist may consider actually employing the algorithm in the real world.
pythonprogramm...
/ sentdex
/ pythonprogramming.net
plus.google.co...
If anyone's getting a "No module named 'Quandl'" error I fixed mine by changing all the 'Quandl's to lowercase.
We were likely researching this at the same time lol.
It's lower-cased now.
Didn't worked for me in the lower case either
lower-cased work for me ...
well if you are using anaconda do a "conda install quandl" which will take care of it. or pip install quandl should do it.
anyone here in 2024
To
Yup
I am.
Never thought Snowden would turn to UA-cam tutorials to support himself.
Amazing video btw!
hhhh true
Honestly a compliment Snowden’s a hero
oh hey there
I was thinking the same .He time travelled back after he got an advice from Joe Rogen .lol
He: uses pure python ide for Machne Learning (Expert teaching Machine Learning)
Me: using Jupyter notebook for learning Machine Learning
If you're getting a No module named 'Quandl' error , change all the 'Quandl' to 'quandl' .
Quandl has now changed . So first go create a free account , confirm your mail id and then copy the API key and do this .
import pandas as pd
import quandl
quandl.ApiConfig.api_key = 'your_api_key_here'
df = quandl.get('WIKI/GOOGL')
print(df.head())
Thank you so much! It is really helpful.
thank you so much
Thanks bruh..
Thanks:)
i an having an issue, my computer is saying: ModuleNotFoundError: No module named 'pandas'
"So the equation of a line... well, we'll talk about more down the line." Heh, nice pun man. Made me chuckle
I get so many errors wile typing the same thing it's inhumane lol.
yo sammmeee
@@dudebro4421bro haha no way dude
Don't you want HL_PCT as (high-low)/low
Yep, I derped.
So should i change it accordingly?
Yeah you should :)
Thank god someone else spotted this, I thought I was stupid for not understanding.
Percentage change is (initial - final) / initial
so shouldn't it be (high-low)/high ??
i am 13 years old rn and i am commenting this so i can come back here years later and say- yea, this is where i started
just study dumbass....instead of doing pointless commentary
did u keep up?
I'm 3, same thing
nah man i didnt@@magnify4720
im 1@@Yellow_300
I just don't understand this. I was never familiar with Stock Market and stuff like that. I can see what you are doing, but I don't know *why* you are doing.
Easiest thing you can do or best thing you can do is get yourself familiar with it, atleast to the point where what he's saying makes sense...but at the end of the day it's just an example, as long as you understand the concepts
I suspect that for line defining 'HL_PCT', you meant to subtract/divide with 'Adj. Low' rather than 'Adj. Close' to get volatility. Looking forward to your vids! You do great tutorials.
+John Dupuy You suspect right! Watched it probably 3 times before publishing, then saw my mistake, d'oh! Luckily, this is just meant to be a super simple example of creating features. That is an interesting feature, though definitely not what was intended. Hopefully you are better than me at the process, and nice catch :P
+John Dupuy thought of the same thing! When I saw the comment I had my little HA! moment haha :). Great video btw +sentdex thanks for putting out all this content!
+John Dupuy Thanks.
+LefkiosG Same :D
sry for my dumb question , why this subtract/divide process is done ? This is dealing with stock calculation? I'm really confusing and getting nervous in finding df[HL_PCT] & df[PCT_Change] :(
df1 = quandl.get_table('WIKI/PRICES')
df = df1[['date','adj_open','adj_high','adj_low','adj_close','adj_volume']]
df.index.names = [None]
df=df.set_index('date')
change the df like the above code for those who are using the new google data. the indexes are not date any more that's the reason the following part doesn't work.
last_date = df.iloc[-1].name
#print (df.index[-1],last_date)
last_unix = last_date.timestamp()
one_day = 86400
next_unix = last_unix + one_day
Roya Norouzi Thanks for your notes. Saved me from having to go through the docs. Cheers.
You now have to use a lowercase 'q' for Quandl:
import quandl
df = quandl.get('WIKI/GOOGL')
Absolutely right. Otherwise, an error came up and you waste a lot of your time juggling around why this is not working?
Than you so much my brother :)...only for u i can make it
I wish I had seen this soon, but after a while I stumbled upon this solution after trying to get quandl into anaconda and seeing it was lowercase.
thanks
Guyzz for following this tutorial in 2020 you need to have an account in Quandl and must confirm account to access data by API key
This code works in 2020✔✔✔
import pandas as pd
import quandl
quandl.ApiConfig.api_key = 'Your Api key'
df = quandl.get('WIKI/GOOGL')
print(df.head())
thx bro it helps a lot
Much appreciated... thanks!
Still works in 2021! Thank you so much!
@@vanessadragonborn1287 ❤️❤️
WIKI/GOOGLE is not available anymore. Can;t find a similar data set to follow this tutorial.
Same. Looking for an alternative. If I find it, I'll post a link here.
It is 'WIKI/GOOGL'
Quandl seem to have changed their simple price call to free data to an API key and request model plus a signup is needed even for free data.
heres what I got when set an account up on Quandl in terms of the python code for a script...
quandl.ApiConfig.api_key = 'xxxxxxxxxxxxxx''
data = quandl.get_table('WIKI/PRICES')
This sort of makes your tutorial, as far as I can see, stop dead at the point of reference to Quandl.get('WIKI/GOOGL') as this just don't work no more.
although to be fair they do still say here that you can just call the prices in, as you outlined, without any need for a key -> www.quandl.com/tools/python
Quandl.get('WIKI/GOOGL') still works without an API key, the thing is that on the quandl website now WIKI/GOOGL, which wasn't in table format, redirects to WIKI/PRICES. Their site seems buggy to me. I started with WIKI/PRICES and API key as well because I'm not like the mindless people who just imitate what they see on a video. Here is another technical link:
github.com/quandl/quandl-python/blob/master/FOR_ANALYSTS.md
Seems like the first stock request is for free. You can do it anonymously. Then the second request of different stock, they may ask you for API keys. You can register a quandl account for free and you will be given you a permanent API key as long as you confirm your email address.
quandl website is completely different now I cannot find anything their
Tbh it’s not hard to use. Okay to use the wiki you don’t have to go website. Ngl the website is mess. The trick is add quote marks a / then stock ticker symbol. Just search company ticker symbol use google. It gets data for that company. Example df= qunadl.get(‘WIKI/F’). Note don’t forget to import quandl and print(df)
@@machinacheater7725 Thank you dude
For those who have installed Python 3.6 and having problems installing sklearn .. save some time and check this link
stackoverflow.com/questions/42360126/error-while-trying-to-install-sklearn-from-pycharm-arrayobject-h-cannot-be-abs
ubuntu/ linux users
if you got any error installing quandl, just type pip3 install quandl. worked to me!
Thank you!!!
Thanks bro , save my life
geniusss! thanks!
its also an improvement for mac users
whats the difference between pip and pip3?
import pandas as pandas
import quandl
df = quandl.get("WIKI/GOOGL")
print(df.head())
#quandl should be lowercase
3 April 2019
I'm happy that people like you exist . In Indian engineering colleges we dont have courses on ML in the computer science stream . Thanks a lot !
well youre studying in the wrong college buddy! VIT has it!
Yeah just pay 16 lakhs. They will teach you ML and provide jobs for spreadsheet and excel lmao.
@@siddhantsingh3392 So true LMAO
@@pranavsebastianlouis may be he misinterpreted Madrasa as engineering college
What are you talking about
My mind was blown when you said "y = mx + b", when my entire life I have been using "c" not "b". Great video though :)
Wow, really?
I use c also. Makes it easier to remember the other formula y-b=m(x-a)
I was taught y = kx + n
Blaž Štempelj
It really doesn't matter. c, n, b as long as we all know what each other mean its the same formula!:)
Totally agree. It's just interesting to see different notations :)
No idea if you catch it in a future video, but I noticed a bug in your feature definition. Your "volatility" measurement should be calculated with high and low, but you are using high and close. I may misunderstand the feature, but you even labeled it HL_PCT.
Yes it should be high minus low!! In his official website he has mentioned this correction.
What's his official website?
pythonprogramming.net/
Ahhh, so that's what got me confused.
I did not understand that part of the video, but I get it now.
Thank you
Are we really having a stock exchange classes here :p
Hello. I’m trying to follow along and it seems Quandle has changed a lot since the last time you used it. For starters, nothing’s free anymore. Found this out after downloading the whole series …
does he use quandl for all other vidoes? im just gettin started rn
real
7:32 : In creating features, shouldn't High-Low percentage be 'High - Low' instead of 'High - Close'. Just an observation.
I noticed that too I don't know if he did it on purpose or not
@@rohanaggarwal8718 9 Months in, almost impressive!
I think so too!
@@ishitashah220 Is that correct or should that be high-low?
Quandl is quandl (small case) for me
+Nick R. Feller Thanks for sharing, wonder why this is different for some people. Are you using a different version of Python?
+sentdex python 3.5
+sentdex I'm on a Mac using 2.7 and it is lowercase too
The same for me. Using Debian 8 and python 2.7.9
It's also lower case for me. Using 3.5.1 on OSX.
I was able to access wiki/googl via python but couldnt find it anywhere on the actual website for the life of me
same haha
the quandl site changed a bit since the video was recorded, if you try to use `WIKI/GOOGL`, you receive an error '(Status 403) (Quandl Error QEPx04) You do not have permission to view vendor datatable 'WIKI/GOOGL'. Please subscribe to this database to get access to the data.'
to fix, just use `quandl.get_table('WIKI/PRICES', ticker='GOOGL')`
also, note that the index is different than in the video. In the latest version in Quandl, the index is NOT the date. In order to make the index the date, like in the video do:
df.set_index(['date'], inplace = True, drop=True)
then everything in the original example will work without having to modify anything else
df = Quandl.get('WIKI/GOOGL')
print(df.head)
works for me and shows the Dataframe.
weird... I'll try again.
thanks for trying!
import quandl as Quandl
df = Quandl.get('WIKI/GOOGL')
print(df)
--
Trying out your code leads to an error which says that an API key is needed
I'm using Quandl 3.1.0 without providing an API key, with python 3.4.3 on PyCharm / macOS
I'm having the same problem. Let me know if you could solve the issue
I was able to confirm that this does work for me. Using python 2.7.11 Very strange why it wasn't working for me before
im also getting this error.......``` File ".\Main.py", line 8, in
mydf["HL_PCT"] = (mydf["Adj. High"] - mydf["Adj. Close"]) / mydf["Adj. Close"] * 100.0
TypeError: list indices must be integers or slices, not str```
i have the same issue. did you ever solve it?
please try enclosing Adj. High,Adj. Close and HL_PCT in a single quotation mark..not double
Your videos are awesome, but i,m a beginner and i think you should make simple regression videos with simple data, thanks
I thought that's what this was :O
If you are using jupyter lab or jupyter notebook, in addition to installing using pip install quandl, open the anaconda cmd, and install conda install quandl to import to jupyter.
You're the best. Thank you!
thank you
another change is if you use pycharm, Instead of using print(df.head), you should use print(df.head()) with parentheses.
Hi! I am loving your videos!
There is a small typo on this one. The HL_PCT should have the Adj. Low instead of Adj. Close in the formula. Nothing much, but I thought I should say it so that someone who might want to use the code gets it right.
Please keep up with the excellent content! S2
Could you also tell me that what does this mean "(high - low)/low" .Is this any formula or something else .Please help
@@surbhirohilla3804 use chatgpt , its gonna help you
@@surbhirohilla3804 In this example, he is trying to get percentage difference in the high and low values. This is just the percentage difference formula.
Alteast have some ads. It really feels bad to watch this videos for free dude!
"I just feel light there's a disconnect between your version and mine"
If any one like me is facing quandl error in ubuntu os, here is solution for that stackoverflow.com/questions/38427430/error-installing-quandl-with-pip-on-ubuntu-14-04
also with pip install sklearn, errors. im going to try using anaconda -it ships with stable sklearn
MikeDrones Yeah another solution could be anaconda... also it is important that some python packages could be 64bit versions also so if solution did not work for someone, so google that same package for 64bit version..
THIS WORKS!!!! Enabling Quandl was huge pain, untill now.
Sweet
Thanx it works to me now :D
Quandl changed :( I can't find abundant free databases on it anymore. I guess I will just walk into a den of lions. And also add what lines of code I am graced with based on my malnourished quandl sample.
try setting an account on Quantopian. You get access to a ton of stock databases
I think quandl is still giving some free datasets not exactly what it is shown in tutorial though.
KeyError: "None of [Index(['adj.open', 'adj.high', 'adj.low', 'adj.close', 'adj.volume'], dtype='object')] are in the [columns]" .........after running this line help me ...df1 =df1[['adj.open','adj.high','adj.low','adj.close','adj.volume',]]
Do you got the answer?
Got through this tutorial. Felt super confused at first but I watched it again and it made a lot more sense.
FOR LINUX USERS IN 2018:
You definitely wanna use a virtual environment for this one, easiest way is to get anaconda
1. download from anaconda.org, then cd ~/Downloads and run with 'bash Anaconda-......' to install anaconda. after that open new console to activate
2. Create a virtual environment using python3.5 (like sentdex) and install pip with: conda create --name python=3.5 pip
3. You can now check your environments with 'conda env list'. Activate your env with 'conda activate (to deactivate simple 'conda deactivate'
4. now that you are in your virtual environment 'python --version' should yield 3.5...
5. Now you can simply your (from command prompt) 'pip install sklearn/quandl/pandas' and all should go smooth (pandas is now included in quandl)
Hope this helped.... took me 2 hours.... :/
Hi! Should I start with this tutorial series just after watching the first 15 videos of python with no prior knowledge in math topics related to machine learing?? Or should I do something else first? I would highly appreciate if someone could help me out.
I also have the same question, I suppose now you know what do I do?
anyone have quandl data link ??
just started learning and I must say, amazing explanations man. Thanks a lot!
Great to hear! Best wishes on your ML journey!
what is df exactly? a dictionary of lists, but how you are using it, i mean by this - df[['xyz', 'abc']]
brackets in brackets
you cant use keys list to redefine dictionary
ok i get it now ..if u are having trouble downloading the dataset just use api key statement like this quandl.ApiConfig.api_key = "your api key after creating free acc " and then use quandl.get statement it works !
Thanks, this solution worked.
u the man siddhesh!
If anyone is getting "ValueError: check_hostname requires server_hostname" error just install this:
pip install urllib3==1.25.8
Where to get the dataset?
hello! I've been stucked with this error, please somebody can help me, thanks
line 6 in df['HL_PCT'] = (df['Adj. High'] - df['Adj. Close']) / df['Adj. Close'] * 100.0
TypeError: list indices must be integers or slices, not str
I'm working with python 3.5.2 and have exactly the same code.
I' m stucked too with this error, are you get the answer thanks
Hello!, I just worked it out. In this line "df = df[["Adj. Open", "Adj. High", "Adj. Low", "Adj. Close", "Adj. Volume",]]", I didn't put " df[[Adj. Open... " instead I had " [[Adj. Open ...".
import pandas as pd
import quandl
df = quandl.get("WIKI/GOOGL")
df = df[["Adj. Open", "Adj. High", "Adj. Low", "Adj. Close", "Adj. Volume",]]
df["HL_PCT"] = (df["Adj. High"] - df["Adj. Low"])/df["Adj. Close"] * 100.0
df["PCT_change"] = (df["Adj. Close"] - df["Adj. Open"])/df["Adj. Open"]* 100.0
df = df[["Adj. Close", "HL_PCT", "PCT_change", "Adj. Volume",]]
print(df.head())
I'm not sure what you changed and when I tried your code it didn't run. maybe because of the commas you have after your column selections. I don't know. But I definitely have the same error.
Also, why are we using double brackets? [[ "column names" ]]
"
import pandas as pd
import quandl
df = quandl.get('WIKI/GOOGL')
df = [['Adj. Open','Adj. High','Adj. Low','Adj. Close','Adj. Volume']]
df['HL_PCT'] = (df['Adj. High'] - df['Adj. Close']) / df['Adj. Close'] *100
df['PCT_change'] = (df['Adj. Close'] - df['Adj. Open']) / df['Adj. Open'] *100
df = df[['Adj. Close','HL_PCT','PCT_change','Adj. Volume']]
print(df.head())
"
Thanks it worked for me!!
@@dermorgendanach93 thx bro
ive not been able to load the 'WIKI/GOOGL' data set it keeps telling me:
' LimitExceededError: (Status 429) (Quandl Error QELx01) You have exceeded the anonymous user limit of 50 calls per day. To make more calls today, please register for a free Quandl account and then include your API key with your requests.
'
i even tried waiting for 24hrs, but the same result
ok i get it now ..if u are having trouble downloading the dataset just use api key statement like this quandl.ApiConfig.api_key = "your api key after creating free acc " and then use quandl.get statement it works !
"I really should be studying for finals"
UA-cam: Let me recommend you this video =)
It will be better if we pratice in an open source instead of using quandl.
Why did you use df[[]] instead of df[] when declaring df but not use df[['HL_PCT']].I know it is a multidimensional list .But why use it
If you want to access multiple columns use df[[]], here inner [] is the list of columns you want to access.
@@parveenk5710 thanks for the explanation!
Why high minus low percent is taken as high minus close divided by close and not high minus low divided by low?
typo mistake from his side, it will be (high-low)/low
Any alternative for quandl? No free data available for Stocks now.
Data is still available just use
df = quandl.get('WIKI/GOOGL')
df = quandl.get('WIKI/GOOGL') doesn't work for me?
me too .. it's give me this Eroor "quandl.errors.quandl_error.LimitExceededError: (Status 429) (Quandl Error QELx01) You have exceeded the anonymous user limit of 50 calls per day. To make more calls today, please register for a free Quandl account and then include your API key with your requests.
"
so try another database .. this one worked for me 'NSE/OIL'
Mohamed, the error you got is because you were using it without an account and you made more than 50 calls as sentdex explained..(hope this cleared it for someone)
Can't find wiki google data set help..!
anyone where to get this dataset?
I pip installed Quandl but when I typed "df = Quandl('')", and it turns out that Quandl is undefined, why is that?
u need to import it with lowercase
If anyone is getting keyError then use "Adj. Close" instead of "Adj.Close "(Use space)
thanks actual so helpful, idk where i went wrong until this
I got the same thing.
i think there is mistake in the column HL_PCT, you have by mistake taken adj_high - adj_close instead of adj_high - adj_low
on printing df.head() after modifying the data frame i am getting error as :
AttributeError: 'list' object has no attribute 'head'
are you sure you did df = pd.read_csv('filename')? If so, try without the parenthesis
sorry i mean df = Quandl.get
@@lucasphillips2177 i didnt find the data sets he uses can you share the link of data set
I opened a repository for everyone who wants to go through the code step-by-step with some small explanations :)
github.com/jousefm/Machine-Learning-Sentdex
Twitter: twitter.com/Jousefm2
Instagram: instagram.com/jousefmrd/
-----
A sub would be great as well! Harrison motivated me to create me own Python series starting soon on my channel :)
for someone who did not show all columns, put: pd.set_option('display.max_rows', 500)
pd.set_option('display.max_columns', 500)
when I paste following lines I get syntax error saying invalid character in identifier
pd.set_option('display.max_rows', 500)
pd.set_option('display.max_columns', 500)
Am i doing something wrong? thanks
cảm ơn bạn nhé
@@GoldDiggersTV did you install or install right version? check all again
i have 3 colums in data shortVolume , shortexemptVolume,TotalVolume
when i tried to get percentage of short this is error im getting
Issue:
df['pct_short'] = df['ShortVolume'] / df['TotalVloume'] * 100.0
TypeError: list indices must be integers or slices, not str
I am trying to print the head after getting data from quandl. But do not know why ,but get api stucks and does not return. Please help
for folks having trouble, you'll need to create an account on quandl to use the data. Once you create an account you get apikey. Run this command before importing data quandl.ApiConfig.api_key = "YOURAPIKEY"
Thank you . IT really helped loading data
Can you describe regression in a simple way and also what you did in this tutorial?
Is it just simple data collection?
Bro if u learned plz teach me. I didnt understand anything. Wtf are they Man
For those getting the error of need to use an API_Key for use...
quandl.ApiConfig.api_key = 'you_api_key_from_quandl'
How do I get my API_Key I already registered. Please share your code. thanks
i think i am too stupid for this .....
@Neil Lunavat Pour some hot coffee on your computer.. Issue resolved
I used NSE:YESBANK with authtoken
It is giving me error like
File "C:\Python27\lib\site-packages\pandas\core\indexing.py", line 1230, in _convert_to_indexer
raise KeyError('%s not in index' % objarr[mask])
KeyError: "['Adj. Open' 'Adj. High' 'Adj. Low' 'Adj. Close' 'Adj. Volume'] not in index"
Getting the same issue.How did u solve it?
Ohhh... Wow You Finally Installed Windows 10..!!!! :)
HL column should be (high-low)/low * 100 right?
You got the data set for 2004 ... How would i get the latest data set as of now i want to get the data set for 2016
at the top of the console it is 2004, if you scroll to the bottom, it is the current date. Sorry for the 10 months late :)
oh and also get rid of df.head() and write df.head instead, not sure why it works but it does
Hello buddy...Awesome video..
I have python 3.6 installed & PIP 19.0.2
Could you please share the link to install sklearn, quandl & pandas...Thank you
Bhai please please help me how to install these.
@@dumbstuff2809 should I turn on my internet while for this?
you also need to pip install NumPy and pip install SciPy first.
Hi Guys, in regards to the Quandl module problem I think it is much easier to use Conda:
1- Install Anaconda. (search how- very easy)
2- use conda install quandl.
3- on my device I used (import Quandl) and it worked.
And Thank you Harrison all the best... :)
tnks man
Thank you for the excellent series, man. Truly doing some great work here.
I did install every thing and I get on Mac Please help
RESTART: /Users/mac/Documents/python/Untitled.py ==============
Traceback (most recent call last):
File "/Users/mac/Documents/python/Untitled.py", line 1, in
import pandas
ModuleNotFoundError: No module named 'pandas'
Make sure that you are running sudo pip3 install pandas OR sudo pip install pandas (if you are using python 2.7)
Dont know if you have figured out the solution yet. But if you are using a Mac, I recommend installing Anaconda distribution. It makes everything much easier.
guys , I have LimitExceededError, could you please send a download file, so I won't get that error, thanks
can you add some q&a session with lot of noob questions?
my question i am in love with maths and programming and the concept of machine learning overwhelms me but i dont know where to start or what path to follow
what path did you follow? I'm in the same boat
KeyError: “['blah'] not in index”
Just change the attribute name to the correct one.
Type print(df.head()) and you can see the correct names.
You're welcome :)
you just saved my life
Hi, I don't speak english very well and yesterday I wached your tutorial of basic python just those 30 first . My question: why did you use two brackets in def =def[[]] ? Please :'D
this means a matrix in Python. There are several columns in this Data Frame(df), it's a 2-dimensional matrix.
If you're wondering how the `df[[ ... ]]` syntax works mechanically, it's overloading __getitem__ for pandas.DataFrame. In other words, Pandas uses "operator overloading" to allow the brackets to be used in new ways. You can overload the accessor (brackets) on your own classes by giving them a __getitem__ method. Another way to view this is that `df[ ["a", "b" ] ]` is the same as `df.__getitem__(["a","b"])`. Hopefully that wasn't too confusing. For more info, check out operator overloading in Python. thepythonguru.com/python-operator-overloading/
Thank You James !!
i thought the same ! in python, u cant have mutable data as key dict. what he does, can only be achieved overloading the "__getitem__" method;
Whenever I try installing sci-kit learn I just get an error. I can install other modules bot not that one please help
same, you got a fix?
Aaargh! Can't wait for more tutorials.
For those of you who have issues with importing Quandl try the following : Have only one interpreter installed on your machine (let's say Anaconda) and use lowercase 'q' (quandl). Hope it helps! Good luck!
Anyone with the latest version of Quandl will have this issue. Somewhere since I did these videos, they changed it. Jerks :P
Is there anyway you could update this content? I'm only 3 minutes in and everything is different now.
I can't find the dataset that you used probably because it's 5 years old , so can anyone help me?
"TypeError: list indices must be integers or slices, not str"
Error is being shown to me,
please help!
me too....
which location of the video did you get this error?
df = df[['Adj. Close','HL_PCT','PCT_change','Adj. Volume']]
Did you type the above correctly?
I got this error when I forgot to type df before the brackets on the right side of the assignment statement: df = df[[...]]
Can you please redo this video a lot has changed apparently and I get soooo many errors from the beginning of the tutorial
Thanks for these ML tutorials. Anaconda Python from Continuum Analytics has all popular Python libraries - SciKit, pandas, NumPy, SciPy...etc
+Kimani Wa Ndirangu This is the most important piece of tutorial which must be told. Installing packages one by one takes forever.
Also, not to forget about the installation errors.
Just to add : Latest Anaconda doesnt have quandl so that needs to be installed to the Anaconda package list. CHeers!
You are writing all these commands in the default python IDE but can I also execute them in PyCharm? I installed all the pip modules after a lot of effort due to errors.
Yes you can use pycharm
@@sentdex Thank you. I plan to try and learn ml on my own because I think it's very interesting and your tutorials help a lot.
Can't fetch data from 'WIKI/GOOGL'
Following error shown:
------------------------------------
Exception has occurred: QuandlError
(Status 403) Something went wrong. Please try again. If you continue to have problems, please contact us at connect@quandl.com.
download this google stock dataset: www.kaggle.com/gunhee/amdgoogle/version/3
replace the quandl with this:
df = pd.read_csv('GOOGL.csv')
you'll ahve to work with unadjusted values though
why the output is looking different from that in the video . will i be able to follow him? thanks :)
not working , any idea why?
@@Alan_christo the dataset he had uses adjusted values so the columns he has say "adj. Close " and the like. This one just has the regular so you use "Close". Just drop adj. Everywhere he uses it
If youre getting a 'ModuleNotFoundError: No module named 'Quandl'' error try importing quandl (no capital) :)
df = df[['Adj. Open', 'Adj. High', 'Adj. Low', 'Adj. Close', 'Adj. Volume']] is this syntax specific to quandl? Or pandas?
Both
if any of you guys having this error (list indices must be integers or slices, not str) make sure you didnt type this like me .
The Wrong ONE !!
df = [['Adj. Open','Adj. High','Adj. Low','Adj. Close','Adj. Volume']]
The True ONE!
df = df[['Adj. Open','Adj. High','Adj. Low','Adj. Close','Adj. Volume']]
i have written the same code and run it. it shows the following errors
-------------------------------------------------------------------------------------------------------------
C:\Python27\python.exe "C:/Users/pavan kumar kota/PycharmProjects/python learning programs/machine learning/mc1.py"
Traceback (most recent call last):
File "C:/Users/pavan kumar kota/PycharmProjects/python learning programs/machine learning/mc1.py", line 2, in
import quandl
File "C:\Python27\lib\site-packages\quandl\__init__.py", line 7, in
from .model.database import Database
File "C:\Python27\lib\site-packages\quandl\model\database.py", line 18, in
import quandl.model.dataset
File "C:\Python27\lib\site-packages\quandl\model\dataset.py", line 5, in
from .data import Data
File "C:\Python27\lib\site-packages\quandl\model\data.py", line 1, in
from quandl.operations.data_list import DataListOperation
File "C:\Python27\lib\site-packages\quandl\operations\data_list.py", line 1, in
from quandl.model.data_list import DataList
File "C:\Python27\lib\site-packages\quandl\model\data_list.py", line 2, in
from .data_mixin import DataMixin
File "C:\Python27\lib\site-packages\quandl\model\data_mixin.py", line 1, in
import pandas as pd
File "C:\Python27\lib\site-packages\pandas\__init__.py", line 26, in
from pandas._libs import (hashtable as _hashtable,
File "C:\Python27\lib\site-packages\pandas\_libs\__init__.py", line 3, in
from .tslib import iNaT, NaT, Timestamp, Timedelta, OutOfBoundsDatetime
File "pandas\_libs\src
umpy.pxd", line 157, in init pandas._libs.tslib (pandas\_libs\tslib.c:117317)
ValueError: numpy.dtype has the wrong size, try recompiling. Expected 52, got 56
Process finished with exit code 1
how to solve the below code? I got a error message.
with 'DeprecationWarning: Function make_blobs is deprecated; Please import make_blobs directly from scikit-learn
warnings.warn(msg, category=DeprecationWarning)'
from IPython.display import display
import pandas as pd
import mglearn
import numpy as np
import scipy as sp
import matplotlib.pyplot as plt
# 데이터셋을 만듭니다
X, y = mglearn.datasets.make_forge()
# 산점도를 그립니다
mglearn.discrete_scatter(X[:, 0], X[:, 1], y)
plt.legend(["클래스 0", "클래스 1"], loc=4)
plt.xlabel("첫 번째 특성")
plt.ylabel("두 번째 특성")
print("X.shape: {}".format(X.shape))
print("mglearn version: {}".format(mglearn.__version__))
X, y = mglearn.datasets.make_wave(n_samples=40)
plt.plot(X, y, 'o')
plt.ylim(-3, 3)
plt.xlabel("특성")
plt.ylabel("타깃")
In case anyone watching this video is stuck, here is my code that works with Python 3.7.0:
import pandas as pd
import quandl
#Gets the data from Quandl about the history of GOOGL from 8/19/2004 - 8/25/2004
df = quandl.get("WIKI/GOOGL", authtoken="", start_date="2004-08-19",
end_date="2004-08-25")
#Edits the Dataframe(DF) to include the columns from the dataset
df = df[["Adj. Open","Adj. High","Adj. Low","Adj. Close","Adj. Volume",]]
#Creates two new columns showing: Change in the high/low and change in the open/close
df["HL_PCT"] = (df["Adj. High"] - df["Adj. Low"]) / df["Adj. Low"] * 100.0
df["CHANGE_PCT"] = (df["Adj. Close"] - df["Adj. Open"]) / df["Adj. Open"] * 100.0
#Changes DF again to show only the columns we want to see, then prints it
df = df[["Adj. High","Adj. Low","HL_PCT","Adj. Open","Adj. Close","CHANGE_PCT",]]
print(df.head())
You need to create a Quandl account now and include your authtoken in your get() request. I changed a couple things that I thought made it show the data a little better but this works. Hope this helps.
hello everyone :)
anybody knows if the whole tutorial "Machine Learning with Python" from our amazing Sentdex is still valid in 2020 (end of the August 2020 to be specific) ?
thaaaaanks
That's why I am here (srry for bad english)
To avoid "limit exceeded" messages when using quandl, I broke down and signed up for a free account to get my own api key. You can call this out in your code with the directive:
```
import quandl
quandl.ApiConfig,api_key = "xxxxxxxxx"
```
If you store it as an os environment variable (say, QUANDL_API_KEY), you won't have to put your key in every script. Instead, just add the lines:
```
import os
import quandl
quandl.ApiConfig.api_key = os.environ['QUANDL_API_KEY']
```
On Linux or Mac, you can add the following line to your .profile (or .bash_profile, depending on your distro):
```
export QUANDL_API_KEY="xxxxxxxxx"
```
On Windows you'll need to open "Advanced System Settings" and click "Environment Variables" to add a new User variable for your Windows user.
I have install python and necessary package in Ubuntu 17.04
My code as below:
import pandas as pd
import quandl
df = quandl.get('WIKI/GOOGL')
print(df.head())
Error below:
:~/Documents/Python$ ./ML1.py
import-im6.q16: unable to open image `pd': Permission denied @ error/blob.c/OpenBlob/2701.
./ML1.py: line 4: syntax error near unexpected token `('
./ML1.py: line 4: `df = quandl.get('WIKI/GOOGL')
As soon as run the program, my cursor turn into snipping tool cursor. Any clue on this?