We Keep Coding

How to normalize complex JSON structure with 4 levels of nested arrays?

As I'm fairly new to python I've tried various ways based on answers found here but can't manage to normalize my json file.
As I checked in Postman it has 4 levels of nested arrays. For suppliers I want to expand all the levels of data.The problem I have is with the score_card subtree for which I want to pull out all risk_groups , then name,risk_score and all riskstogether with name,risk_score and indicators with the deepest level containing again name and risk_score.
and I'm not sure whether it is possible in one go based on ambiguous names between levels.
Below I'm sharing a reproducible example:
data = {"suppliers": [
{
"id": 1260391,
"name": "2712270 MANITOBA LTD",
"duns": "rm-071-7291",
"erp_number": "15189067;15189067",
"material_group_ids": [
176069
],
"business_unit_ids": [
13728
],
"responsible_user_ids": [
37761,
37760,
37759,
37758,
37757,
37756,
36520,
37587,
36494,
22060,
37742,
36446,
36289
],
"address": {
"address1": "BOX 550 NE 16-20-26",
"address2": None,
"zip_code": "R0J 1W0",
"city": "RUSSELL",
"state": None,
"country_code": "ca",
"latitude": 50.7731176,
"longitude": -101.2862461
},
"score_card": {
"risk_score": 26.13,
"risk_groups": [
{
"name": "Viability",
"risk_score": 43.33,
"risks": [
{
"name": "Financial stability supplier",
"risk_score": None,
"indicators": [
{
"name": "Acquisitions",
"risk_score": None
}
]
}
]
},
]
}
}
]
}
And here is how it should look:
expected = [[1260391,'2712270 MANITOBA LTD','rm-071-7291',176069,13728,[37761,37760,
37759,37758,37757,37756,36520,37587,36494,22060,37742,36446,36289],
'BOX 550 NE 16-20-26','None','R0J 1W0','RUSSELL','None','ca',50.7731176,
-101.2862461,26.13,'Viability',43.33,'Financial stability supplier','None',
'Acquisitions','None']]
df = pd.DataFrame(expected,columns=['id','name','duns','material_groups_ids',
'business_unit_ids','responsible_user_ids',
'address.address1','address.address2','address.zip_code',
'address.city','address.state','address.country_code','address.latitude',
'address.longitude','risk_score','risk_groups.name',
'risk_groups.risk_score','risk_groups.risks.name',
'risk_groups.risks.risk_score','risk_groups.risks.indicators.name',
'risk_groups.risks.indicators.risk_score'])
What I tried since last 3 days is to use json_normalize:
example = pd.json_normalize(data['suppliers'],
record_path=['score_card','risk_groups','risks',
'indicators'],meta = ['id','name','duns','erp_number','material_group_ids',
'business_unit_ids','responsible_user_ids','address',['score_card','risk_score']],record_prefix='_')
but when I specify the fields for record_path which seems to be a vital param here it always return this prompt:
TypeError: {'name': 'Acquisitions', 'risk_score': None} has non list value Acquisitions for path name. Must be list or null.
What I want to get a a table with columns as presented below:
So it seems that python doesn't know how to treat most granular level which contains null values.
I've tried approach from here: Use json_normalize to normalize json with nested arrays
but unsuccessfully.
I would like to get the same view which I was able to 'unpack' in Power BI but this way of having the data is not satisfying me.
All risk indicators are the components of risks which are part of risk groups as You can see on attached pictures.
Going to the most granular level it goes this way: Risk groups -> Risks -> Indicators so I would like to unpack it all.
Any help more than desperately needed. Kindly thank You for any support here.

Format an f-string for each dataframe object

Requirement
My requirement is to have a Python code extract some records from a database, format and upload a formatted JSON to a sink.
Planned approach
1. Create JSON-like templates for each record. E.g.
json_template_str = '{{
"type": "section",
"fields": [
{{
"type": "mrkdwn",
"text": "Today *{total_val}* customers saved {percent_derived}%."
}}
]
}}'
2. Extract records from DB to a dataframe.
3. Loop over dataframe and replace the {var} variables in bulk using something like .format(**locals()))
Question
I haven't worked with dataframes before.
What would be the best way to accomplish Step 3 ? Currently I am
3.1 Looping over the dataframe objects 1 by 1 for i, df_row in df.iterrows():
3.2 Assigning
total_val= df_row['total_val']
percent_derived= df_row['percent_derived']
3.3 In the loop format and add str to a list block.append(json.loads(json_template_str.format(**locals()))
I was trying to use the assign() method in dataframe but was not able to figure out a way to use like a lambda function to create a new column with my expected value that I can use.
As a novice in pandas, I feel there might be a more efficient way to do this (which may even involve changing the JSON template string - which I can totally do). Will be great to hear thoughts and ideas.
Thanks for your time.

I would not write a JSON string by hand, but rather create a corresponding python object and then use the json library to convert it into a string. With this in mind, you could try the following:
import copy
import pandas as pd
# some sample data
df = pd.DataFrame({
'total_val': [100, 200, 300],
'percent_derived': [12.4, 5.2, 6.5]
})
# template dictionary for a single block
json_template = {
"type": "section",
"fields": [
{"type": "mrkdwn",
"text": "Today *{total_val:.0f}* customers saved {percent_derived:.1f}%."
}
]
}
# a function that will insert data from each row
# of the dataframe into a block
def format_data(row):
json_t = copy.deepcopy(json_template)
text_t = json_t["fields"][0]["text"]
json_t["fields"][0]["text"] = text_t.format(
total_val=row['total_val'], percent_derived=row['percent_derived'])
return json_t
# create a list of blocks
result = df.agg(format_data, axis=1).tolist()
The resulting list looks as follows, and can be converted into a JSON string if needed:
[{
'type': 'section',
'fields': [{
'type': 'mrkdwn',
'text': 'Today *100* customers saved 12.4%.'
}]
}, {
'type': 'section',
'fields': [{
'type': 'mrkdwn',
'text': 'Today *200* customers saved 5.2%.'
}]
}, {
'type': 'section',
'fields': [{
'type': 'mrkdwn',
'text': 'Today *300* customers saved 6.5%.'
}]
}]

How to convert dynamic nested json into csv?

I have some dynamically generated nested json that I want to convert to a CSV file using python. I am trying to use pandas for this. My question is - is there a way to use this and flatten the json data to put in the csv without knowing the json keys that need flattened in advance? An example of my data is this:
{
"reports": [
{
"name": "report_1",
"details": {
"id": "123",
"more info": "zyx",
"people": [
"person1",
"person2"
]
}
},
{
"name": "report_2",
"details": {
"id": "123",
"more info": "zyx",
"actions": [
"action1",
"action2"
]
}
}
]
}
More nested json objects can be dynamically generated in the "details" section that I do not know about in advance but need to be represented in their own cell in the csv.
For the above example, I'd want the csv to look something like this:
Name, Id, More Info, People_1, People_2, Actions_1, Actions_2
report_1, 123, zxy, person1, person2, ,
report_2, 123, zxy , , , action1 , action2
Here's the code I have:
data = json.loads('{"reports": [{"name": "report_1","details": {"id": "123","more info": "zyx","people": ["person1","person2"]}},{"name": "report_2","details": {"id": "123","more info": "zyx","actions": ["action1","action2"]}}]}')
df = pd.json_normalize(data['reports'])
df.to_csv("test.csv")
And here is the outcome currently:
,name,details.id,details.more info,details.people,details.actions
0,report_1,123,zyx,"['person1', 'person2']",
1,report_2,123,zyx,,"['action1', 'action2']"

I think what your are lookig for is:
https://pandas.pydata.org/docs/reference/api/pandas.json_normalize.html

If using pandas doesn't work for you, here's the more canonical Python way of doing it.
You're trying to write out a CSV file, and that implicitly means you must write out a header containing all the keys.
The constraint that you don't know the keys in advance means you can't do this in a single pass.
def convert_record_to_flat_dict(record):
# You need to figure out exactly how you want to do this; everything
# should be strings.
record.update(record.pop('details'))
return record
header = {}
rows = [[]] # Leave the header row blank for now.
csv_out = csv.writer(buffer)
for record in report:
record = convert_record_to_flat_dict(record)
for key in record.keys():
if key not in header:
header[key] = len(header)
rows[0].append(key)
row = [''] * len(header)
for key, index in header.items():
row[index] = record.get(key, '')
rows.append(row)
# And you can go back to ensure all rows have the same number of keys:
for row in rows:
row.extend([''] * (len(row) - len(header)))
Now you have a list of lists that's ready to be sent to csv.csvwriter() or the like.
If memory is an issue, another technique is to write out a temporary file and then reprocess it once you know the header.

Filtering pandas dataframe by date to count views for timeline of programs

I need to count viewers by program for a streaming channel from a json logfile.
I identify the programs by their starttimes, such as:
So far I have two Dataframes like this:
The first one contains all the timestamps from the logfile
viewers_from_log = pd.read_json('sqllog.json', encoding='UTF-8')
# Convert date string to pandas datetime object:
viewers_from_log['time'] = pd.to_datetime(viewers_from_log['time'])
Source JSON file:
[
{
"logid": 191605,
"time": "0:00:17"
},
{
"logid": 191607,
"time": "0:00:26"
},
{
"logid": 191611,
"time": "0:01:20"
}
]
The second contains the starting times and titles of the programs
programs_start_time = pd.DataFrame.from_dict('programs.json', orient='index')
Source JSON file:
{
"2019-05-29": [
{
"title": "\"Amiről a kövek mesélnek\"",
"startTime_dt": "2019-05-29T00:00:40Z"
},
{
"title": "Koffer - Kedvcsináló Kul(t)túrák Külföldön",
"startTime_dt": "2019-05-29T00:22:44Z"
},
{
"title": "Gubancok",
"startTime_dt": "2019-05-29T00:48:08Z"
}
]
}
So what I need to do is to count the entries / program in the log file and link them to the program titles.

My approach is to slice log data for each date range from program data and get the shape. Next add column for program data with results:
import pandas as pd
# setup test data
log_data = {'Time': ['2019-05-30 00:00:26', '2019-05-30 00:00:50', '2019-05-30 00:05:50','2019-05-30 00:23:26']}
log_data = pd.DataFrame(data=log_data)
program_data = {'Time': ['2019-05-30 00:00:00', '2019-05-30 00:22:44'],
'Program': ['Program 1', 'Program 2']}
program_data = pd.DataFrame(data=program_data)
counts = []
for index, row in program_data.iterrows():
# get counts on selected range
try:
log_range = log_data[(log_data['Time'] > program_data.loc[index].values[0]) & (log_data['Time'] < program_data.loc[index+1].values[0])]
counts.append(log_range.shape[0])
except:
log_range = log_data[log_data['Time'] > program_data.loc[index].values[0]]
counts.append(log_range.shape[0])
# add aditional column with collected counts
program_data['Counts'] = counts
Output:
Time Program Counts
0 2019-05-30 00:00:00 Program 1 3
1 2019-05-30 00:22:44 Program 2 1

A working (but maybe a little quick and dirty) method:
Use the .shift(-1) method on the timestamp column of programs_start_time dataframe, to get an additional column with a name date_end indicating the timestamp of end for each TV program.
Then for each example_timestamp in the log file, you can query the TV programs dataframe like this: df[(df['date_start']=<example_timestamp) & (df['date_end']>example_timestamp)] (make sure you substitute df with your dataframe's name: programs_start_time) which will give you exactly one dataframe row and extract from it the name of the TV programm.
Hope this helps!

Solution with histogram, using numpy:
import pandas as pd
import numpy as np
df_p = pd.DataFrame([
{
"title": "\"Amiről a kövek mesélnek\"",
"startTime_dt": "2019-05-29T00:00:40Z"
},
{
"title": "Koffer - Kedvcsináló Kul(t)túrák Külföldön",
"startTime_dt": "2019-05-29T00:22:44Z"
},
{
"title": "Gubancok",
"startTime_dt": "2019-05-29T00:48:08Z"
}
])
df_v = pd.DataFrame([
{
"logid": 191605,
"time": "2019-05-29 0:00:17"
},
{
"logid": 191607,
"time": "2019-05-29 0:00:26"
},
{
"logid": 191611,
"time": "2019-05-29 0:01:20"
}
])
df_p.startTime_dt = pd.to_datetime(df_p.startTime_dt)
df_v.time = pd.to_datetime(df_v.time)
# here's part where I convert datetime to timestamp in seconds - astype(int) casts it to nanoseconds, hence there's // 10**9
programmes_start = df_p.startTime_dt.astype(int).values // 10**9
viewings_starts = df_v.time.astype(int).values // 10**9
# make bins for histogram
# add zero to the beginning of the array
# add value that is time an hour after the start of the last given programme to the end of the array
programmes_start = np.pad(programmes_start, (1, 1), mode='constant', constant_values=(0, programmes_start.max()+3600))
histogram = np.histogram(viewings_starts, bins=programmes_start)
print(histogram[0]
# prints [2 1 0 0]
Interpretation: there were 2 log entries before 'Amiről a kövek mesélnek' started, 1 log entry between starts of 'Amiről a kövek mesélnek' and 'Koffer - Kedvcsináló Kul(t)túrák Külföldön', 0 log entries between starts of 'Koffer - Kedvcsináló Kul(t)túrák Külföldön' and 'Gubancok' and 0 entries after start od 'Gubancok'. Which, looking at the data you provided, seems correct :) Hope this helps.
NOTE: I assume, that you have the date of the viewings. You don't have them in the example log file, but they appear in the screenshot - so I assumed that you can compute/get them somehow and added them by hand to the input dict.

Series of JSON objects to dataframe conversion

I have downloaded a sample dataset from here that is a series of JSON objects. According to the website, each JSON object looks like below
{
"id": "4cd223df721b722b1c40689caa52932a41fcc223",
"title": "Knowledge-rich, computer-assisted composition of Chinese couplets",
"paperAbstract": "Recent research effort in poem composition has focused on the use of automatic language generation...",
"entities": [
"Conformance testing",
"Natural language generation",
"Natural language processing",
"Parallel computing",
"Stochastic grammar",
"Web application"
],
"s2Url": "https://semanticscholar.org/paper/4cd223df721b722b1c40689caa52932a41fcc223",
"s2PdfUrl": "",
"pdfUrls": [
"https://doi.org/10.1093/llc/fqu052"
],
"authors": [
{
"name": "John Lee",
"ids": [
"3362353"
]
},
"..."
],
"inCitations": [
"c789e333fdbb963883a0b5c96c648bf36b8cd242"
],
"outCitations": [
"abe213ed63c426a089bdf4329597137751dbb3a0",
"..."
],
"year": 2016,
"venue": "DSH",
"journalName": "DSH",
"journalVolume": "31",
"journalPages": "152-163",
"sources": [
"DBLP"
],
"doi": "10.1093/llc/fqu052",
"doiUrl": "https://doi.org/10.1093/llc/fqu052",
"pmid": ""
}
Eventually I need to work with the paperAbsrtract section only. I am loading this to a pandas dataframe like below
filename = "sample-S2-records"
df = pd.read_json(filename, lines=True)
df.head()
This shows all the doi and doiUrl column empty.
Also if I only select abstract column and check out the head, I see 2 of the 5 rows empty
abstract = df['paperAbstract']
abstract.head()
0
1 The search for new administrators in complex s...
2 The human N-formyl peptide receptor (FPR) is a...
3 Serum CA 19-9 (2-3 sialyl Le(a)) is a marker o...
4
Name: paperAbstract, dtype: object
Looks like the way I have created the dataframe is not the right approach. I am pretty confident they don't have any missing column.
What am I missing? Any suggestion?

I looked into a sample of your data, and I think you're getting correct results. If we were to parse the JSON by hand:
import json
filename = "sample-S2-records"
with open(filename, 'r') as f:
d = [json.loads(x) for x in f]
Then inspect the list of dictionaries, here's what we see:
>>> d[0]['paperAbstract']
''
So it really looks like the first lines paperAbstract field is empty.
P.S.: I think that the question needs to be closed, I doubt it's going to help anyone else