We Keep Coding

Missing value replacemnet using mode in pandas in subgroup of a group

Having a data set as below.Here I need to group the subset in column and fill the missing values using mode method.Here specifically needs to fill the missing value of Tom from UK. So I need to group the TOM from Uk, and in that group the most repeating value needs to be added to the missing value.
Below fig shows how i need to do the group by.From the below matrix i need to replace all the Nan values using mode.
the desired output:
attaching the dataset
Name location Value
Tom USA 20
Tom UK Nan
Tom USA Nan
Tom UK 20
Jack India Nan
Nihal Africa 30
Tom UK Nan
Tom UK 20
Tom UK 30
Tom UK 20
Tom UK 30
Sam UK 30
Sam UK 30

try:
df = df\
.set_index(['Name', 'location'])\
.fillna(
df[df.Name.eq('Tom') & df.location.eq('UK')]\
.groupby(['Name', 'location'])\
.agg(pd.Series.mode)\
.to_dict()
)\
.reset_index()
Output:
Name location Value
0 Tom USA 20
1 Tom UK 20
2 Tom USA NaN
3 Tom UK 20
4 Jack India NaN
5 Nihal Africa 30
6 Tom UK 20
7 Tom UK 20
8 Tom UK 30
9 Tom UK 20
10 Tom UK 30
11 Sam UK 30
12 Sam UK 30

Update columns with duplicate values from the DataFrame in Pandas

I have a data set which has values for different columns as different entries with first name to identify the respective columns.
For instance James's gender is in first row and James's age is in 5th row.
DataFrame
df1=
Index
First Name
Age
Gender
Weight in lb
Height in cm
0
James
Male
1
John
175
2
Patricia
23
5
James
22
4
James
185
5
John
29
6
John
176
I am trying to make it combined into one DataFrame as below
df1=
Index
First Name
Age
Gender
Weight
Height
0
James
22
Male
185
1
John
29
175
176
2
Patricia
23
I tried to do groupby but it is not working.

Assuming NaN in the empty cells, you can use groupby.first:
df.groupby('First Name', as_index=False).first()
output:
First Name Age Gender Weight in lb Height in cm
0 James 22.0 Male 185.0 NaN
1 John 29.0 None 175.0 176.0
2 Patricia 23.0 None NaN NaN

Copy contents from one Dataframe to another based on column values in Pandas

Two seperate similar DataFrames with different lengths
df2=
Index
First Name
Age
Gender
Weight
0
James
25
Male
155
1
John
27
Male
175
2
Patricia
23
Female
135
3
Mary
22
Female
125
4
Martin
30
Male
185
5
Margaret
29
Female
141
6
Kevin
22
Male
198
df1=
Index
First Name
Age
Gender
Weight
Height
0
James
25
Male
165
5'10
1
John
27
Male
175
5'9
2
Matthew
29
Male
183
6'0
3
Patricia
23
Female
135
5'3
4
Mary
22
Female
125
5'4
5
Rachel
29
Female
123
5'3
6
Jose
20
Male
175
5'11
7
Kevin
22
Male
192
6'2
df2 has some rows which are not in df1 and df1 has some values which are not in df2.
I am comparing df1 against df2. I have calculated the newentries with the following code
newentries = df2.loc[~df2['First Name'].isin(df1['First Name'])]
deletedentries = df1.loc[~df1['First Name'].isin(df2['First Name'])]
where newentries denote the rows/entries that are there in df2 but not in df1; deletedentries denote the rows/entries that are there in df1 but not in df2. The above code works perfectly fine.
I need to copy the height from df1 to df2 when the first names are equal.
df2.loc[df2['First Name'].isin(df1['First Name']),"Height"] = df1.loc[df1['First Name'].isin(df2['First Name']),"Height"]
The above code copies the values however indexing is causing an issue and the values are not copied to the corresponding rows, I tried to promote First Name as the Index but that doesn't solve the issue. Please help me with a solution
Also, I need to calculate the modified values, if the First Name is same, I need to check for modified values; for example in df1, the weight of James is 155 however in df2 the weight is 165, so I need to store the modified weight of James(165) and index(0) in a new dataframe without iteration; the iteration takes a long time because this is a sample of a big dataframe with a lot more rows and columns.
Desired output:
df2=
Index
First Name
Age
Gender
Weight
Height
0
James
25
Male
155
5'10
1
John
27
Male
175
5'9
2
Patricia
23
Female
135
5'3
3
Mary
22
Female
125
5'4
4
Martin
30
Male
185
5
Margaret
29
Female
141
6
Kevin
22
Male
198
6'2
Martin's and Margaret's heights are not there in df1, so their heights are not updated in df2
newentries=
Index
First Name
Age
Gender
Weight
Height
4
Martin
30
Male
185
5
Margaret
29
Female
141
deletedentries=
Index
First Name
Age
Gender
Weight
Height
2
Matthew
29
Male
183
6'0
5
Rachel
29
Male
123
5'3
6
Jose
20
Male
175
5'11
modval=
Index
First Name
Age
Gender
Weight
Height
0
James
165
7
Kevin
192

Building off of Rabinzel's answer:
output = df2.merge(df1, how='left', on='First Name', suffixes=[None, '_old'])
df3 = output[['First Name', 'Age', 'Gender', 'Weight', 'Height']]
cols = df1.columns[1:-1]
modval = pd.DataFrame()
for col in cols:
modval = pd.concat([modval, output[['First Name', col + '_old']][output[col] != output[col + '_old']].dropna()])
modval.rename(columns={col +'_old':col}, inplace=True)
newentries = df2[~df2['First Name'].isin(df1['First Name'])]
deletedentries = df1[~df1['First Name'].isin(df2['First Name'])]
print(df3, newentries, deletedentries, modval, sep='\n\n')
Output:
First Name Age Gender Weight Height
0 James 25 Male 155 5'10
1 John 27 Male 175 5'9
2 Patricia 23 Female 135 5'3
3 Mary 22 Female 125 5'4
4 Martin 30 Male 185 NaN
5 Margaret 29 Female 141 NaN
6 Kevin 22 Male 198 6'2
First Name Age Gender Weight
4 Martin 30 Male 185
5 Margaret 29 Female 141
First Name Age Gender Weight Height
2 Matthew 29 Male 183 6'0
5 Rachel 29 Male 123 5'3
6 Jose 20 Male 175 5'11
First Name Age Gender Weight
0 James NaN NaN 165.0
6 Kevin NaN NaN 192.0

for your desired output for df2 you can try this:
desired_df2 = df2.merge(df1[['First Name','Height']], on='First Name', how='left')
#if you want to change the "NaN" values just add ".fillna(fill_value=0)" for e.g 0 after the merge
print(desired_df2)
First Name Age Gender Weight Height
0 James 25 Male 155 5'10
1 John 27 Male 175 5'9
2 Patricia 23 Female 135 5'3
3 Mary 22 Female 125 5'4
4 Martin 30 Male 185 NaN
5 Margaret 29 Female 141 NaN
6 Kevin 22 Male 198 6'2
new and deleted entries is already right. for the moment I'm a bit stuck how to get the modval dataframe. I'll update my answer if I get a solution.

pandas long_to_wide method based on input rows

I am having issue deleting nulls. My input dataframe
name no city tr1_0 tr2_0 tr3_0 tr1_1 tr2_1 tr3_1 tr1_2 tr2_2 tr3_2
John 11 edi boa 51 110 cof 52 220
Rick 12 new cof 61 100 dcu 61 750
Mat t1 nyc
my desired output
name no city tr1 tr3 tr2
0 John 11 edi boa 110 51
1 John 11 edi cof 220 52
2 Rick 12 new cof 100 61
3 Rick 12 new dcu 750 61
4 Matt 13 wil nan nan nan
i used below code
df1 = pd.read_fwf(inputFileName, widths=widths, names=names, dtype=str, index_col=False )
feature_models = [col for col in df1.columns if re.match("tr[0-9]_[0-9]",col) is not None]
features = list(set([ re.sub("_[0-9]","",feature_model) for feature_model in feature_models]))
ub("_[0-9]","",feature_model) for feature_model in feature_models]))
df1 = pd.wide_to_long(df1,i=['name', 'no',
df1 = pd.wide_to_long(df1,i=['name', 'no', 'city',],j='ModelID',stubnames=features,sep="_")
my current output as below. row 2 doesn't make any sense in my use case so i don't want to generate that row at all. if there is no trailer i only want 1 row which is good (row 6). if there are 2 trailers,i only want 2 rows but its giving me 3 rows. (row 2 and row 5 are extra). i tried using dropna but its not working. Also in my case its printing as nan not NaN.
name no city tr1 tr3 tr2
0 John 11 edi boa 110 51 .
1 John 11 edi cof 220 52 .
2 John 11 edi nan nan nan .
3 Rick 12 new cof 100 61 .
4 Rick 12 new dcu 750 61 .
5 Rick 12 new nan nan nan .
6 Matt 13 wil nan nan nan .

You can use this alternative solution with split and stack:
df1 = df1.set_index(['name', 'no', 'city'])
df1.columns = df1.columns.str.split('_', expand=True)
df1 = df1.stack(1, dropna=False).reset_index(level=3, drop=True)
mask = df1.index.duplicated() & df1.isnull().all(axis=1)
df1 = df1[~mask].reset_index()
print (df1)
name no city tr1 tr2 tr3
0 John 11 edi boa 51.0 110.0
1 John 11 edi cof 52.0 220.0
2 Rick 12 new cof 61.0 100.0
3 Rick 12 new dcu 61.0 750.0
4 Mat t1 nyc NaN NaN NaN
With your solution:
df1 = pd.wide_to_long(df1,i=['name', 'no', 'city'],j='ModelID',stubnames=features,sep="_")
For remove NaNs with duplicated MultiIndex values is possible use filtering by boolean indexing:
#remove counting level
df1 = df1.reset_index(level=3, drop=True)
mask = df1.index.duplicated() & df1.isnull().all(axis=1)
df1 = df1[~mask].reset_index()
Details:
Check dupes by Index.duplicated:
print (df1.index.duplicated())
[False True False True False True]
Then check missing values by DataFrame.all for check all Trues per rows:
print (df1.isnull().all(axis=1))
name no city
John 11 edi False
edi False
Rick 12 new False
new False
Mat t1 nyc True
nyc True
dtype: bool
Chain by & for bitwise AND:
mask = df1.index.duplicated() & df1.isnull().all(axis=1)
print (mask)
name no city
John 11 edi False
edi False
Rick 12 new False
new False
Mat t1 nyc False
nyc True
dtype: bool
Invert boolean mask by ~:
print (~mask)
name no city
John 11 edi True
edi True
Rick 12 new True
new True
Mat t1 nyc True
nyc False
dtype: bool

Pandas: Dict of data frames to unbalanced Panel

I have a dictionary of DataFrame objects:
dictDF={0:df0,1:df1,2:df2}
Each DataFrame df0,df1,df2 represents a table in a specific date of time, where the first column identifies (like social security number) a person and the other columns are characteristics of this person such as
DataFrame df0
id Name Age Gender Job Income
10 Daniel 40 Male Scientist 100
5 Anna 39 Female Doctor 250
DataFrame df1
id Name Age Gender Job Income
67 Guto 35 Male Engineer 100
7 Anna 39 Female Doctor 300
9 Melissa 26 Female Student 36
DataFrame df2
id Name Age Gender Job Income
77 Patricia 30 Female Dentist 300
9 Melissa 27 Female Dentist 250
Note that the id (social security number) identifies exactly the person. For instance, the same "Melissa" arises in two different DataFrames. However, there are two different "Annas".
In these dataFrames the number of people and the people vary over time. Some people is represented in all dates and others are represented only in a specific date of time.
Is there a simple way to transform this dictionary of data frames in an (unbalanced) Panel object, where the ids arise in all dates and if the data a given id is not available it is replaced by NaN?
Off course, I can do that making a list of all ids and then checking in each date if a given id is represented. If it is represented, then I copy the data. Otherwise, I just write NaN.
I wonder if there an easy way using pandas tools.

I would recommend using a MultiIndex instead of a Panel.
First, add the period to each dataframe:
for n, df in dictDF.iteritems():
df['period'] = n
Then concatenate into a big dataframe:
big_df = pd.concat([df for df in dictDF.itervalues()], ignore_index=True)
Now set your index to period and id and you are guaranteed to have a unique index:
>>> big_df.set_index(['period', 'id'])
Name Age Gender Job Income
period id
0 10 Daniel 40 Male Scientist 100
5 Anna 39 Female Doctor 250
1 67 Guto 35 Male Engineer 100
7 Anna 39 Female Doctor 300
9 Melissa 26 Female Student 36
2 77 Patricia 30 Female Dentist 300
9 Melissa 27 Female Dentist 250
You can also reverse that order:
>>> big_df.set_index(['id', 'period']).sort_index()
Name Age Gender Job Income
id period
5 0 Anna 39 Female Doctor 250
7 1 Anna 39 Female Doctor 300
9 1 Melissa 26 Female Student 36
2 Melissa 27 Female Dentist 250
10 0 Daniel 40 Male Scientist 100
67 1 Guto 35 Male Engineer 100
77 2 Patricia 30 Female Dentist 300
You can even unstack the data quite easily:
big_df.set_index(['id', 'period'])[['Income']].unstack('period')
Income
period 0 1 2
id
5 250 NaN NaN
7 NaN 300 NaN
9 NaN 36 250
10 100 NaN NaN
67 NaN 100 NaN
77 NaN NaN 300