30- Exploratory Data Analysis (EDA) with Python (Day-14)

[@Codanics]

Please like and share the video, comment below if you have any question?

[@talhanisarmughal]

12:42 Brute Force technique to find the unique values of all columns def func(att): return titanic[att].unique() for i in titanic.columns: print(i,func(i))

[@muhammadawon8164]

Your every video is Gold!! Thank you

[@FireyLeo007]

57:15 # for quick comparison between fare and fare_log ks_clean.boxplot('fare') ks_clean.boxplot('fare_log')

[@faizasandhu_]

I was waiting for your video💥 And was already watching yesterday's zoom session evern after attending meeting, ky kahin kuch miss na ho gya ho😀 PS: I also have a class at 9am But don't wanna miss these classes Baba Aamar_ GREAT👍

[@engrismatullahkhan3828]

What an awesome work you are doing Sir .. Absolutely brilliant . Stay blessed .

[@daniabatool7931]

12:42 to find unique values from multiple columns (np.unique(kashti[['sex', 'who']].values))

[@syedabdulwahabshah3280]

for int and float value it's not working , put (age,survived)and than check TypeError: '

[@muhammadammarmohsin107]

50:35 Correlation is a statistical term describing the degree to which two variables move in coordination with one another. If the two variables move in the same direction, then those variables are said to have a positive correlation. If they move in opposite directions, then they have a negative correlation.

[@taqi_haider]

12:40 Assignment : to find the uniques values in multiple columns pd.concat([ks['survived'],ks['deck'],ks['Age'],ks['fare'],ks['class'],ks['survived']]).unique()

[@Fayez-tk7ph]

Timestamp: 57:12 Assignment # 02: The log transformation has made the graph clearer and easier to understand. In other words, we can say the data is less skewed.

[@ikrambashir8604]

12:00. Finding the code for unique values in multiple columns (boat['sex'].append(boat['alive'])).unique() output: array(['male', 'female', 'no', 'yes'], dtype=object)

[@NasirJumani]

@0:05 What a start baba G. New looks!

[@Codanics]

Ap bor na ho jao is liay

[@NasirJumani]

@@Codanics :D

[@FireyLeo007]

11:25 # To pull out unique values from the kashti dataset "ks", the best way to approach is by "for-loop" function kdata = ks.columns # defining a new string to pull out the index names for kda in kdata: # starting for -loop print(ks[kda].unique()) # print out the values for individual unique column print()

[@attaullah4998]

Hazrat Behtareen ho gya hai. Learning and Enjoying

[@syedkamranhaider9467]

12:23 To get unique values for all the columns in pandas dataframe following line of code is used: print(ks.apply(lambda col: col.unique())) Explanation lambda is used to create single line function without name ''col'' is iterator which goes through every column in "ks" dataframe to find the unique values.

[@asfandyarsaeed6402]

11:24 Assgingment solution: Q: find out the unique of all columns soultion_Theory : Use pandas.DataFrame.values on the ks to convert the DataFrame to a numpy.array. Call numpy.array.ravel() to flatten the numpy.array. Use pandas.unique(column_values) to get the unique values of the previous result column_values. Soultion_Code: col1=ks[:].values.ravel() col2=pd.unique(col1) print(col2) ## here we got with Baba G result set

[@asfandyarsaeed6402]

Incase you want to see 3 column the do this col1=ks[['sex','who','class']].values.ravel() col2=pd.unique(col1) print(col2) output:: ['male' 'man' 'Third' 'female' 'woman' 'First' 'child' 'Second']

[@khawarabbas815]

Ma sha Allah very good way and easy way of teaching

[@goharrahman5611]

11:25 Assignment 1 col1=ks[:].values.ravel() col2=pd.unique(col1) print(col2) 43:27 the plot of mean can be plot by the following code ks_clean.groupby(['sex', 'class']).mean().plot.bar()

[@Zeeshan18382]

EDA ko itna asan kar diya apny kaya bat hai Dr sahab very nice

[@muhammadsair8594]

43:10 Group by plot code # Group by function to see data in groups and deploy plot ks_clean.groupby(['sex','class']).mean().plot.bar()

[@mugheezAhmed]

12:30 Assignment Here we use function of 'for' Unique = Ks columns For k in unique: Print(Ks[k].unique()) Print() :) Output: All unique values of column through one line of code

[@AbdulHannan-dg6dl]

11:23 np.unique(boat[['sex', 'who','class']].values)) # Selective data more than one column for col in boat: print(boat[col].unique()) # Getting all data 30:56 new_boat[new_boat['age']

[@syedabdulwahabshah3280]

for int and float value it's not working , put (age,survived)and than check TypeError: '

[@aajizattari]

11:40 np.unique (ks[['pclass', 'parch','alone']].values) Output: 0, True, 2, 3, 4, 5, 6]

[@salmanirshad8356]

11:27 #assignment ks[['survived', 'pclass', 'sex', 'age', 'sibsp', 'parch', 'fare', 'embarked', 'class', 'who', 'adult_male', 'deck', 'embark_town', 'alive', 'alone']].nunique()

[@MujeebJan]

12:44 pd.concat([ks['survived'],ks['who'],ks['deck'],ks['sex'],ks['class'],ks['adult_male'],ks['pclass'],ks['parch']]).unique() output: array([0, 1, 'man', 'woman', 'child', nan, 'C', 'E', 'G', 'D', 'A', 'B', 'F', 'male', 'female', 'Third', 'First', 'Second', 3, 2, 5, 4, 6], dtype=object)

[@natureisrelaxing1053]

12:09 To find the unique values of more than one columns pd.concat([ks['sex'],ks['who'],ks['survived']]).unique()

[@atiyashaheen6624]

57:20 log se large number ko small numbers mein kr skte hain jis se uski graphical representation convenient ho jati hy

[@zeeshanrafeeque9135]

unique values of 2 or more columns in single line of code: pd.concat([ks['sex'],ks['pclass'],ks['who']]).unique()

[@muhammadzeeshanafzal1996]

Thank you bro your answer helped me to solve this query.. JazakAllah

[@AyeshaFatima-zf5yn]

Best video till now

[@arslanakbar7783]

12:20 Assignment # Get unique elements in multiple columns i.e. Sex & Class & Who (pd.concat((ks_clean['sex'],(ks_clean['class']),(ks_clean['who'])))).unique()

[@doptopgaming9569]

excellent job sir respect

[@mumairali766]

11:33 we can check the unique values in the two or more columns by using the [ [ ] ] f

[@AounHussain]

12:41 Assignment #To find unique values from two columns ks['who'].append(ks['adult_male']).unique() Output: array(['man', 'woman', 'child', True, False], dtype=object)

[@AounHussain]

33:58 It should be written as ks_clean=ks_clean.loc[ks_clean['age']

[@AounHussain]

57:14 The Boxplot with 'fare_log' has very few outliers whereas that with 'fare' has many outliers

[@ymentertainment2349]

#12:55 Time pause Assignment bt in my case and bt mean boat and boat mean kashti bt[["survived", "pclass", 'sex', 'age', 'sibsp', 'parch', 'fare', 'embarked', 'class', 'who', 'adult_male', 'deck', 'embark_town', 'alive', 'alone']].nunique()

[@data-science-tutors]

11:23 here we use for loop kd = ks.columns for k in kd: print(ks[k].unique()) print()

[@FireyLeo007]

24.30 WORD OF CAUTION The data's stats (describe) has changed due to removal of the missing data. One needs to be very cautious as it has impact on your findings i.e. refined vs actual For example by removing the approx 177 rows of age, the survived rate went up which is not a true representation So one needs to be very vigilant and cautious as what you are removing from the dataset and its impact on your final report/findings.

[@Codanics]

Exactly

[@ahsanzafar4921]

while removing outliers for age column, when data filtered with age < 66, it completely removes the outliers. So, i think the technique is to change values until you get no outlier to the value at which it gives outlier, now choose a value just below it, all ouliers will be removed.

[@AllInOnekh]

@11:23 k=ks.columns for i in k: e=ks[i].nunique() print(e)

[@faryalineurope]

In 2nd step of cleaning data, where you are trying to plot a bar graph using groupby function. If we put () after "bar" then the code can work. ks_clean.groupby([ 'sex' , 'class' ]). mean(). plot. bar()

[@sardarabdullahkhawar8487]

12:30 for i in ks.columns: print(ks[i].unique())

[@muhammadsair8594]

40:32 & 50:06 To See and save plots need to install extensions i)VSCode Plot and ii)Jupyter Notebook Renderers _without these extensions in Dark Theme x and y-axis values are not visible as they are also black and users are not allowed to save and view in zoom the graphs._

[@mugheezAhmed]

Thank you sir n yes it's working now

[@Fayez-tk7ph]

Timestamp: 11:16 Assignment - ks[['survived', 'sex', 'parch']].nunique()

[@sks8102]

12:09. Finding the code for unique values of more than one columns

[@aasimnawaz]

if anyone wanna change to light theme, kindly change from (File>preferences>color theme) short key is , ctrl+k Ctrl+T

[@mmateenalihashmi7440]

(12:40) for col in ks: print(ks[col].unique())

[@manazzaayub8430]

Stamptime 12.27 Assignment How to find unique value of all columns pd.concat([ks['survived'],ks['adult_male'],ks['alive'],ks['sex'],ks['class'],ks['who']]).unique() array([0, 1, 'no', 'yes', 'male', 'female', 'Third', 'First', 'Second', 'man', 'woman', 'child'], dtype=object)

[@SajidSpeaks5]

12:41 # multiple unique values ks[["age", "class", "alive", "alone"]].nunique() #output age 88 class 3 alive 2 alone 2 dtype: int64

[@Bobthemagnificent]

Great video

[@mugheezAhmed]

40:45 I am unable to save and complete view output images which extension need to be installed? Plz guide anyone thank you in advance :)

[@usman1694]

31:09 BABAJI agr ks_clean ki age ka mean check kia tha to ks ki age qw update ki ? ks_clean ki hi age update hogi na??

[@resreng]

For unique values of multiple coloms pd.DataFrame({"values":{col:ks[col].unique() for col in ks}})

[@sweela1]

This one best, I want to learn its logic .

[@e-learnwithsoniairfan5226]

36:18 891-712=179, ( where as 177 are age null values, 2 null values are of embarked and 2 are of embark town) ad embark town is string data so only 2 is deducted.

[@tahreemzaheer8994]

at 42:39 if the code will be ks_clean.groupby(['sex','class']).mean().plot.bar() it would plot the bar plot. the code was giving an error because of missing ()

[@muhammadzohaib8517]

df=pd.DataFrame(ks) unique_values=df.nunique() print(unique_values) Print all Unique value of all columns

[@faizasandhu_]

12:14 ks[["who","survived","age","fare"]].nunique() Output who 3 survived 2 age 88 fare 248 dtype: int64

[@humairrazzaq7871]

12:52 ks.loc([:,['who','gender']) i hope, it will work mene ye kiya huwa hai warna iloc se get kr skte hain

[@muhammadabubakar8930]

#12:17 Assignment ks[['sex','who','age']].nunique() Output: sex 2 who 3 age 83 dtype: int64

[@danishnawaz1625]

12:32 pd.unique(ks[['deck', 'sex']].values.ravel('K'))

[@nida7825]

#12:26 Assignment pd.concat([ks['survived'],ks['who'],ks["deck"],ks["sex"],ks["class"]]).unique()

[@muhammadaamirkhan5666]

love u sir life changing lec

[@mansoorjan9238]

pd.concat([ship['sex'], ship['age'],ship['who']]).unique() for different columns uniqueness.....

[@arehman3304]

Try this one for all coloumn ks[0:].nunique()

[@zainulabidin7013]

11:55 Unique values (ks. ['who'].append(ks['adult_male'])).unique()

[@ymentertainment2349]

#50:03 time pause meray pass yai option hi show ni hota plot save ka ...any one can tell me how to enable this option?

[@irfanKhan-ct8ln]

Many many thanks sir

[@aasimnawaz]

34:31, operators should be defined in parenthesis, ks_clean=ks_clean[ks_clean['age']

[@rahatnaqvi]

Hi I followed same as you mentioned and then ks_clean.head() But doesn't age change to numbers. Any solution please?

[@arslanahmad148]

**Assignment#1 :** ks[['sex','who','age','pclass','deck']].nunique() **Output** sex 2 who 3 age 88 pclass 3 deck 7 dtype: int64

[@ashhadahmed6026]

# 12:43 assignment (solution) pd.unique(ks[['who','sex']].values.ravel())

[@mumtazhamid]

Masha Allah

[@ujala.tasneem2737]

Dark theme was quite Good.

[@usmankhanjadoon]

At 39:50, how much data can we remove is there any limit, any correlation or justification?

[@iqraiqbal3365]

very productive lec

[@umerfarooq757]

Sir one question we late comers can also summit the assignment or not or if we want then how

[@muhammadsair8594]

11:26 Unique Values of multiple columns # find unique values of same type of data columns. This function will not work if data types are different. column_values = ks[["sex",'who','alive']].values unique_values = np.unique([column_values]) unique_values

[@hhamzazaibbhatti1765]

for col in df: print(df[col].unique()) #for all unique values

[@syedabdulwahabshah3280]

whata about varaible from 2 ,3 ,4 or colums plus check it for int and float values

[@kashifraza3339]

27:55 Q: Does normal distribution of Numerical data is necessary for data analysis??

[@Codanics]

Yes i will explain later

[@tahreemzaheer8994]

57:14 the fare_log box plot has less outliers as compared to the fare

[@mugheezAhmed]

Yes exactly I have too outliers with fairlog but less than fare..

[@shahidkamalworld7931]

40:40 option of saving graph and viewing it in separate window is missing for me. How to resolve?

[@izharulhaq3296]

dark theme acha hai sir, I feel comfortable using it.

[@asadtariq6080]

43:22 You missed the function brackets after plot.bar ks.clean.groupby(['sex', 'class']).mean().plot.bar()

[@shoaibahmeddayo626]

Thanks bro Dr AAmmar

[@InfobyAR]

11:23 np.unique(df[['sex', 'who','class']].values)

[@syedabdulwahabshah3280]

for int and float value it's not working , put (age,survived)and than check TypeError: '

[@ameerabbas8619]

57:41 Sir oper wala box plot x="sex" and y="age" hain y="fare_log" nahi hai es leye out liers show ho rahy hai. baki fare_log ko plot karny se b koch out lier ab b hai. thank you

[@ahmedullah8799]

57:13 the difference is cleaning data from outlierss

[@SureshKumar-dw3xs]

@57:40 Baba G "fare_log" wale plot ko "age" wale plot se compare kr rahe hn, jb k hamen "fare" wale se comparison krni hai

[@usmankhanjadoon]

New look...

[@kabeerfcc4153]

Sir video lectures Jo upload kiye Hein wo kafee Hein python seekhnay k liyee?

[@Codanics]

Hanji foran seekh jao jitni jaldi hy ap ko

[@zeeshanahmed430]

12.30 how to get unique values from multiple or all columns, below is the code. uniq_Values = (sh['survived'].append(sh['pclass']).append(sh['sex']).append(sh['age']).append(sh['sibsp']).append(sh['parch']).append(sh['embarked']) .append(sh['class']).append(sh['who']).append(sh['adult_male']).append(sh['deck']).append(sh['embark_town']).append(sh['alive']).append(sh['alone'])).unique() print(uniq_Values)

[@alinaqi8250]

Ammar Bhai please make a video for Netcdf(3d and 4d) datasets which are most important for research students. thank you in advance

[@ikrambashir8604]

thank God I fixed the ks in the start

[@nasirnauman308]

12:35 assignment

[@haseeboffc]

12:42 ks[['survived', 'pclass', 'sex', 'age', 'sibsp', 'parch', 'fare', 'embarked', 'class', 'who', 'adult_male', 'deck', 'embark_town', 'alive', 'alone']].nunique()

[@sunailahtariq8503]

Sir after data cleaning you are comparing survival rate of cleaned and uncleaned data but in actual there was 0 null value in survived column so how null values impact that factor please explain more... And also esa krne se hum data loss nahi kr rae??

[@aasukhan3635]

# Assignment1 ks[['survived', 'pclass', 'sex', 'age', 'sibsp', 'parch', 'fare']].nunique() as these are n numbers so we will use n before using unique option.

[@Lahori369]

Superb

[@AR-Edzone]

Assignment 01 Column _values =ks[['class', 'alnoe ' ]]. Values.ravel() Unique_values=pd.unique(column_values)

[@usman1694]

baba ji aap ny ks_clean ki age update karny k bajaye ks ki age update krdi jis sy aagay masla hua lkn maine ks_clean ki hi age update ki video k sath sath practice krty huwe

[@syedabdulwahabshah3280]

@Codanic while finding unique values when we use string and float or int value ot gives error

[@jafarabbas632]

# unique values for multiple columns select_col = tit[['sex', 'who']].values.ravel() unique_val = pd.unique(select_col) unique_val output: array(['male', 'man', 'female', 'woman', 'child'], dtype=object)

[@hajraanwar8213]

57:12 outliers eliminated after log

[@hassanorakzai8981]

@12:09 ks[['who','class','survived']].nunique()

[@usmankhanjadoon]

That's great