Merge all data frame into a single file

I have a list of data frames which looks like this:

df1

Df2

Like that I have many data frames

Merge all into single file and keep only the seller whose latest values are all greater than 0.

Final DF would be like -

Note - In df1 Hari's value was greater than 0 but in df2 it's value is 0 so we have not included in our final DF.

concat the DataFrames in the desired order, then use a pivot_table, getting the last value per Seller and post-filter using loc, gt and all to keep the rows with only positive values:

(pd.concat([df1, df2])
   # sort by Month-year (use %b-%d if Month-day)
   .sort_values(by='Date', key=lambda x: pd.to_datetime(x, format='%b-%y'))
    # reshape to wide form, keeping only the last value
   .pivot_table(index='Seller', columns='Date', values='Vaule', aggfunc='last')
    # keep only rows with all positive values
   .loc[lambda d: d.gt(0).all(axis=1)]
   .reset_index().rename_axis(columns=None)
)

Output:

    Seller  Jan-02  Jan-03
0      Mat    14.0    14.0
1    Messi     1.0    10.0
2       Om    12.0    14.0
3  Ronaldo     7.0    17.0

intermediate before the loc filtering step:

Date     Jan-02  Jan-03
Seller                 
Cat         0.0     NaN
Hari        2.0     0.0
John        0.0     0.0
Mat        14.0    14.0  # those rows only
Messi       1.0    10.0  # have positive values
Om         12.0    14.0  # keep them
Ron         NaN     0.0
Ronaldo     7.0    17.0  #

Another possible solution:

(pd.concat([df1, df2])
 .drop_duplicates('Seller',keep='last')
 .loc[lambda x: x['Vaule'].ne(0)]
 .dropna())

Output:

    Seller  Value    Date
1       Om   12.0  Jan-02
3      Mat   14.0  Jan-02
5    Messi   10.0  Jan-02
7  Ronaldo    7.0  Jan-02

To answer the requirement:

(pd.concat([df1, df2])
 .drop_duplicates('Seller',keep='last')
 .loc[lambda x: x['Vaule'].ne(0)]
 .dropna()
 .merge(df1, on='Seller', suffixes=('_jan03', '_jan02'))
 .loc[:, ['Seller', 'Vaule_jan02', 'Vaule_jan03']])

Output:

    Seller  Value_jan02  Value_jan03
0       Om           12         14.0
1      Mat           14         14.0
2    Messi            1         10.0
3  Ronaldo            7         17.0

A more general approach, which is based on reduce combined with merge:

from functools import reduce

list_df = [df1, df2]

(reduce(lambda x, y:  
    pd.merge(x, y, on='Seller', suffixes=['', f"_{y['Date'][0]}"]), list_df)
 .rename({'Value': f"Value_{df1['Date'][0]}"}, axis=1)
 .loc[lambda x: x.iloc[:,-2].gt(0), lambda x: ~x.columns.str.startswith('Date')])

Use concat with remove sellers if Values are less like 0, remove duplicates and aggregate sum:

df = pd.concat([df1, Df2])

df = (df[~df['Seller'].isin(df.loc[df['Vaule'].le(0), 'Seller'])]
        .drop_duplicates()
        .groupby('Seller', as_index=False)['Vaule']
        .sum()
        )
print (df)
    Seller  Vaule
0      Mat   14.0
1    Messi   10.0
2       Om   12.0
3  Ronaldo    7.0