Need suggestions for a better approach to handle missing keys in a function operating on dictionaries

The question is that I am seeking the best practice to solve my task.
I have the function that I described below (simple example for the explanation of the problem)

def create_additional_keys(data: dict):
  data['l_t_j'] = 1 if data['a'] in [27, 11, 33] else 0
  data['b_n_j'] = 1 if data['b'] in [29, 1, 27] else 0
  data['p_t_q'] = 'ck' if data['c'] == '' else data['c']
  data['m_k_z'] = 'd12' if data['d'] in ['d1', 'd2'] else 'other'
  data['y_s_n'] = data['e1'] * data['e2'] * data['e3']
  data['h_g_p'] = np.log(data['f'])
  ...
  data['s_t_x'] = 1 if data['g'] < 0 else data['g']
  data['c_e_m'] = 1 if data['i'] in [97, 26, 57] else 2 if data['i'] in [98, 27, 58] else 3
  data['s_o_j'] = 1 if data['j'] in [82, 38, 60] else 0
  data['k_s_a'] = data['h'] // 4

The problem is that when I use this function, I always need to ensure that my dictionary has all the keys, but it is not always comfortable.
I often have the majority of keys that I need, but sometimes I do not.
What is the best practice to make a function independent of whether I have these keys or not?

At this point, I have several variants of realization, but I don't like them much and want to realize the best variant.

1. To wrap each code in try-except (as I said, most often, the dictionary has a majority of keys)
   ex.:

try:
  data['l_t_j'] = 1 if data['a'] in [27, 11, 33] else 0
except KeyError:
  pass

2. Before formatting a new key, check first if the key that is needed exists in the dictionary.
   ex.:

if 'a' in data:
  data['l_t_j'] = 1 if data['a'] in [27, 11, 33] else 0

3. To move the lines of code responsible for creating a new key to separate functions and use a loop with try-except structure to iterate through them
   ex.:

formation_l_t_j = lambda data: {"l_t_j": 1 if data["a"] in [27, 11, 33] else 0}
...
formation_k_s_a = lambda data: {"k_s_a": data["h"] // 4}
for function in [formation_l_t_j, ..., formation_k_s_a]:
  try:
    data.update(function(data))
  except KeyError:
    pass

I reviewed your last proposition and implement it in a way that I found cleaner:

data = {'a': 27, 'b': 11}

l = (
    ('l_t_j', lambda: data['a'] in [27, 11, 33]),
    ('b_n_j', lambda: data['b'] in [29, 1, 27]),
    ('p_t_q', lambda: 'ck' if data['c'] == '' else data['c']),
)

for key, compute_value in l:
    try:
        value = compute_value()
    except KeyError:
        continue
        
    data[key] = value

print(data)
# {'a': 27, 'b': 11, 'l_t_j': True, 'b_n_j': False}

It's usually not recommended to "name" a lambda, we usually do a standard function definition instead.

https://peps.python.org/pep-0008/#programming-recommendations

> Always use a def statement instead of an assignment statement that
> binds a lambda expression directly to an identifier:
> python
> # Correct:
> def f(x): return 2*x
>
> python
> # Wrong:
> f = lambda x: 2*x
>
> The first form means that the name of the resulting function object is
> specifically ‘f’ instead of the generic ‘<lambda>’. This is more
> useful for tracebacks and string representations in general. The use
> of the assignment statement eliminates the sole benefit a lambda
> expression can offer over an explicit def statement (i.e. that it can
> be embedded inside a larger expression)

The get() method is the typical way to handle missing keys. If will return the value or None if the key is absent. There is also an optional parameter to use when the key is not there.

To get 1 if present 0 if not or no key

data['l_t_j'] = 1 if data['a'] in [27, 11, 33] else 0

# becomes

data['l_t_j'] = int( data.get('a') in [27, 11, 33] )  

Using or to replace empty/None strings:

data['p_t_q'] = 'ck' if data['c'] == '' else data['c']

# becomes

data['p_t_q'] =  data.get('c') or 'ck'

To select between two values:

data['m_k_z'] = 'd12' if data['d'] in ['d1', 'd2'] else 'other'

# becomes:

data['m_k_z'] = 'd12' if data.get('d') in ['d1', 'd2'] else 'other'

# or this way:

data['m_k_z'] =  'd12' * (data.get('d') in ['d1', 'd2']) or 'other'

Using default parameter for numerical calculations:

data['y_s_n'] = data['e1'] * data['e2'] * data['e3']

# becomes

data['y_s_n'] = data,get('e1',0) * data.get('e2',0) * data.get('e3',0)

Function calls:

data['h_g_p'] = np.log(data['f'])

# becomes 

data['h_g_p'] = np.log(data.get('f',1)) 
# assuming you want zero when key is absent, otherwise ...

data['h_g_p'] = data.get(f) and np.log(data['f']) # None if ansent

default value to chose condition result when key is missing:

data['s_t_x'] = 1 if data['g'] < 0 else data['g']

# becomes

data['s_t_x'] = 1 if data.get('g',-1) < 0 else data['g']

Multiple conditions:

data['c_e_m'] = 1 if data['i'] in [97, 26, 57] else 2 if data['i'] in [98, 27, 58] else 3

#becomes

data['c_e_m'] = {97:1, 26:1, 57:1, 98:2, 27:2, 58:2}.get(data.get('i'),3)