I have one dataframe that I need to filter based on condition receiving in request from UI. request example:

{
"table": "abc",
"condition": "A=98 and C=73 and D='rendom_char'"
} 

dataframe sample:

so suppose if I get condition from UI that "condition": "A=98 and C=73 and D='def'" or "condition": "A=98 and C=73" output should be like :

The issue I'm facing is how to convert that string condition getting from UI into python form so that I can apply filter on datafram.

CodePudding user response：

Use DataFrame.query with replace = to double =:

d = {"condition": "A=98 and C=73 and D='def'"}
print (df.query(d['condition'].replace('=','==')))
    A   B   C    D
2  98  17  73  def
3  98  52  73  def

d = {"condition": "A=85 and C=21"}
print (df.query(d['condition'].replace('=','==')))

    A   B   C    D
4  85  52  21  rst

CodePudding user response：

Alternatively, you can query pandas dataframe in SQL fashion with pandasql:

import pandasql as psql
...

req = {"table": "abc", "condition": "A=98 and C=73 and D='def'"}
q = f"select * from df where {req['condition']}"

print(psql.sqldf(q, locals()))

    A   B   C    D
0  98  17  73  def
1  98  52  73  def