CODEWITHSUNDEEP
pythonpandasjupyter-notebook
OLGJ
OLGJ

Reputation: 432

Compare two column values from different dataframes

I have two dataframes which I've read with pandas. Both contains a Date column and a Stock column, and I want to find out if the corresponding values in those two columns are matching. If they match I would like to update test_version with the corresponding Volume & Price values from unique_values.

I am using Python and Jupyter notebook.

# unique_values
       Index Stock  Date      Volume       Price   Score
0          1   ASO     1   4650600.0   31.139999  0.5719
272      273   GME     1   6218300.0  184.500000  0.9995
403      404   AMC     1  44067000.0   10.200000  0.9995
435      436  TSLA     1  28271800.0  691.619995  0.9686
509      510   AMD     1  29327900.0   81.440002  0.9686
...      ...   ...   ...         ...         ...     ...
11185  11186  AAPL    15  94812300.0  133.110001 -0.9399
11292  11293  BABA    15  12093900.0  229.880005  0.3907
11302  11303  CLOV    15  41659000.0    8.620000  0.9519
11464  11465   NIO    15  71208600.0   36.930000  0.4588
11478  11479  MVIS    15  16808800.0   10.390000  0.9753

[192 rows x 6 columns]

# test_version
   Stock Date  Volume  Price     Score
0    GME    1       1      1  0.194760
1    GME    2       1      1  0.126104
2    GME    3       1      1  0.041961
3    GME    4       1      1  0.039760
4    GME    5       1      1  0.105480
..   ...  ...     ...    ...       ...
10  CLOV   11       1      1       NaN
11  CLOV   12       1      1  0.145852
12  CLOV   13       1      1  0.224382
13  CLOV   14       1      1  0.226059
14  CLOV   15       1      1  0.120781

[210 rows x 5 columns]

I am uncertain if I've approached the problem correctly, but here's what I've tried:

unique_volume.reset_index(drop=True)
test_version.reset_index(drop=True)

test_version['Volume'] = np.where(test_version['Date'] == unique_volume['Date'] and test_version['Stock'] == unique_volume['Stock'], unique_volume['Volume'])


#Output
ValueError: Can only compare identically-labeled Series objects

I'm aspiring to get an output in the form of something like this:

# Desired Output
   Stock Date  Volume  Price     Score
0    GME    1   6218300.0  184.500000  0.194760
..   ...  ...     ...    ...       ...

14  CLOV   15   6218300.0  184.500000  0.120781

[210 rows x 5 columns]

Upvotes: 1

Views: 369

Answers (1)

Shivam Roy
Shivam Roy

Reputation: 2061

If I understood your question correctly, merging (pd.merge : left join)the DataFrames should work for you:

test_version = pd.merge(test_version[['Date', 'Stock']], unique_volume[['Date', 'Stock', 'Volume', 'Price']], on = ['Date', 'Stock'], how = 'left')

Upvotes: 1

Related Questions