code refactoring

[lorenzua02]

Hello, this PR is a refractor of your code. That means it won't change the behaviour of the code, but the code quality will be improved: writing nice code is important to collaborate with other people :)

I suppose you are new to python: for this reason, I am writing here some "mistakes" you've made with the solution & explanation. Hit me up for any question!

---

if status['status'] == 0:
    return True
else:
    return False

This part status['status'] == 0 is a boolean value already! You can simply return it. learn more

return status['status'] == 0

---

IMPORTANT: dictionary as default parameter

Dictionary as default values in functions might not work as you except. This is a serious issue!

If you change the dictionary, it stays like that also in the following calls.

I highly suggest to check out the "Default Parameter Values in Python" link you find on this question on stackoverflow.

---

open() has a context manager: it's suggested to perform file operations using it. In other words, I've replaced your old code...

f = open(...)
# something...
f.close()

...with the same thing, using context manager

with open(...) as f:
    # something... (file will be automatically closed)

---

Some lines could have be inserted in a for loop.

# old
dataframe_copy = ta_sma.calc_ta_sma(dataframe_copy, 5)
dataframe_copy = ta_sma.calc_ta_sma(dataframe_copy, 8)
dataframe_copy = ta_sma.calc_ta_sma(dataframe_copy, 15)
dataframe_copy = ta_sma.calc_ta_sma(dataframe_copy, 20)
dataframe_copy = ta_sma.calc_ta_sma(dataframe_copy, 50)
dataframe_copy = ta_sma.calc_ta_sma(dataframe_copy, 200)

# new
for val in [5, 8, 15, 20, 50, 200]:
    df = ta_sma.calc_ta_sma(df, val)

---

You should comment the meaning of the code, not the actual line! A quick example of comments I've deleted:

# Return updated order dataframe
return updated_order_dataframe

---

About if: Instead of using just the else part, you can reverse the condition, quick example

old:

if something:
    pass
else:
    # code

new:

if not something:
    # code

---

It's really a common practice (even in its documentation) to import pandas as pd and import numpy as np

---

In functions, you should totally avoid nested conditions & returns!
Old code:

if first_condition:
    if second_condition:
        if third_condition:
            return 0
        else:
            raise MyAwesomeException
    else:
        return -2
else:
    return -1

Refractored, more readable, new code:

if not first_condition:
    return -1
   
if not second_condition:
    return -2

if not third_condition:
    raise MyAwesomeException

return 0  # <- Gets here in case it's all good with the conditions above

---

If there is some code at the end that will get excecuted in any scenario, you can move it ouside. Example from your (old) code:

if paper:
    insert_paper...
    return True
else:
    insert_live...
    return True

new:

if paper:
    insert_paper...
else:
    insert_live...
   
return True

---

Tests in python should be like tests/test_trade_capture.py file, please move all the tests (functions) in a class Something(unittest.TestCase)

---

pythonic ways to do things

Old	New
if x == True	if x
if x == None	if is None
x = x * (-1)	x *= -1
if x == "a" or x == "b" or x == "c"	if x in ["a", "b", "c"]
range(0, n)	range(n)

---

Replace

if var == "M1":
    result = "foo"
elif var == "H6":
    result = "bar"
else:
    raise Exception

with

xyz = {"M1": "foo", "H6": "bar"}
# if var not in xyz:
#    raise Exception
result = xyz[var]

(or upgrade to python 3.10 and use the match)

---

You don't have to create SQL string manually to handble database-related stuffs, there are more pythonic ways to do that.

Check PEP8 style guide

There is a lot of duplicated code you can improve, maybe move it in a function and call it every time you need it. One function call is better then the same 5 lines everytime!

test and "real" API_KEYS (& info) should be changed in an .env file or something like that. The script shouldn't even know which is the current enviroment. You could have a PROD_ENV and a TEST_ENV files and just select what env file to use. There are a lot of options tho! .env library

---

I hope everything makes sense to you, feel free to comment on this PR and we will discuss :)

See you soon,

Lorenzo