熊猫：对表2中的表1中的值执行布尔测试

Question

I have a question on Pandas.

I have table1 and table2 with the following values. table2 basically contains a set of random values for each "name":

table1 = pd.DataFrame([{'name': 'A', 'value': 10},
                    {'name': 'B', 'value': 12},
                    {'name': 'C', 'value': 13}])

table2 = pd.DataFrame.from_dict({'name': ['A', 'B', 'C',
                                      'A', 'B', 'C',
                                      'A', 'B', 'C',
                                      'A', 'B', 'C',
                                      'A', 'B', 'C',
                                      'A', 'B', 'C',
                                      'A', 'B', 'C'],
                             'value': [np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30),
                                       np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30),
                                       np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30),
                                       np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30),
                                       np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30),
                                      np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30),
                                      np.random.randint(0,20), np.random.randint(0,30), np.random.randint(0,30)]})

I want to:

1. Create a new column in table2 called "Greater/Less Than"
2. Take each "name" from table1 and check the rows that contain the corresponding name in table2
3. Check whether the value of "A", "B", or "C" in each row in table2 is less or greater than the value of "A", "B", or "C" specified in table1.
4. Return a string of "Less than" or "Greater than" in the new column in table2.

This can be generalized to N number of columns in table1, not just a "name" column. I've googled high and low but I am only able to find basic boolean operators "eg table2[table2[value] <= 100]" but not by using a range of values from another table.

I'm new to Python. Any help would be appreciated! Thank you!

Answer 1

This is one way. Just map values from table1 to table2 , perform the comparison and convert the Boolean to your desired result.

table2['t1_value'] = table2['name'].map(table1.set_index('name')['value'])
table2['Greater/Less Than'] = (table2['value'] < table2['t1_value'])\
                               .map({True: 'Less Than', False: 'Greater Than'})
table2 = table2.drop('t1_value', 1)

#    name  value Greater/Less Than
# 0     A      7         Less Than
# 1     B      3         Less Than
# 2     C     24      Greater Than
# 3     A      9         Less Than
# 4     B     16      Greater Than
# 5     C     17      Greater Than
# 6     A      0         Less Than
# 7     B      5         Less Than
# 8     C     23      Greater Than
# 9     A      4         Less Than
# 10    B     11         Less Than
# 11    C      7         Less Than
# 12    A     11      Greater Than
# 13    B      8         Less Than
# 14    C      9         Less Than
# 15    A      0         Less Than
# 16    B     28      Greater Than
# 17    C     20      Greater Than
# 18    A      8         Less Than
# 19    B     23      Greater Than
# 20    C      2         Less Than

Answer 2

By using np.where

table2['Greater/Less Than']=np.where((table2.set_index('name')-table1.set_index('name')).gt(0),'Greater','Less')
table2
Out[31]: 
   name  value Greater/Less Than
0     A      1              Less
1     B     27              Less
2     C     25              Less
3     A      7              Less
4     B      9           Greater
5     C     20              Less
6     A      3              Less
7     B     25           Greater
8     C     24              Less
9     A      9           Greater
10    B      2              Less
11    C     25           Greater
12    A     17              Less
13    B     16           Greater
14    C      3           Greater
15    A      8           Greater
16    B      1           Greater
17    C     13           Greater
18    A      3              Less
19    B     25              Less
20    C     29           Greater