Metadata object - how to convert to dictionary or Pandas DataFrame?

[andrew-d-moses]

Hi all,

I'm having a difficult time wrapping my head around exactly what the metadata object is, and I'd love to be able to convert it to a more familiar format, such as a dictionary or a Pandas DataFrame. Additionally, I'm passing this to the next Django view via a session variable, so it should be serialized to some degree and converted into a format which can be passed. Any ideas? I'm not very familiar with JSON or BytesIO, for example, but will absolutely continue researching it if it's an option and somebody can get me started in the right direction. Below is a brief example of what I'd like to do:

views.py:

import pandas as pd
import pyreadstat
from django.shortcuts import redirect

def View1(request):
    # read uploaded .sav file from MEDIA_ROOT, validate the ModelForm, grab the path + name of the file (file_name)
    df, meta = pyreadstat.read_sav(filename_path=file_name)
    df_dict = df.to_dict()
    ''' How to convert metadata to dictionary here? '''
    request.session['df_dict'] = df_dict
    request.session['meta_dict'] = meta_dict
    return redirect ('app_name:View2')

def View2(request):
    df_dict = request.session['df_dict']
    meta_dict = request.session['meta_dict']
    # do some things
    # return next view

[MDS-JAnthony]

Morning,
If you haven't already read the docs, this is where I would start.
https://ofajardo.github.io/pyreadstat_documentation/_build/html/index.html

This is a function that I wrote to convert the metadata to json

def metadata_to_json(metadata):
    """
    Function to turn metadata into json object. 
    :param: metadata dataframe to redeuce memory consumption
    :return: json
    """
    metadata_json = {
        "column_names": metadata.column_names,
        "column_labels": metadata.column_labels,
        "column_names_to_labels": metadata.column_names_to_labels,
        "number_columns": metadata.number_columns,
        "variable_value_labels": metadata.variable_value_labels,
        "value_labels": metadata.value_labels,
        "variable_to_label": metadata.variable_to_label,
        "original_variable_types": metadata.original_variable_types,
        "readstat_variable_types": metadata.readstat_variable_types,
        "variable_alignment": metadata.variable_alignment,
        "variable_storage_width": metadata.variable_storage_width,
        "variable_display_width": metadata.variable_display_width,
        "variable_measure": metadata.variable_measure
    }
    return json.dumps(metadata_json)

[ofajardo]

metadata is an ordinary python object, therefore you can easily transform it to a dict using the vars function:

df, meta = pyreadstat.read_sav(filename_path=file_name)
meta_dict = vars(meta)

and with that you can convert it to json:

import json
myjson = json.dumps(vars(meta))

[andrew-d-moses]

Fantastic, thank you for the prompt reply! This is very helpful! Andy Moses

…

On Thu, May 11, 2023, 3:47 AM James Anthony ***@***.***> wrote: Morning, If you haven't already read the docs, this is where I would start. https://ofajardo.github.io/pyreadstat_documentation/_build/html/index.html This is a function that I wrote to convert the metadata to json def metadata_to_json(metadata): """ Function to turn metadata into json object. :param: metadata dataframe to redeuce memory consumption :return: json """ metadata_json = { "column_names": metadata.column_names, "column_labels": metadata.column_labels, "column_names_to_labels": metadata.column_names_to_labels, "number_columns": metadata.number_columns, "variable_value_labels": metadata.variable_value_labels, "value_labels": metadata.value_labels, "variable_to_label": metadata.variable_to_label, "original_variable_types": metadata.original_variable_types, "readstat_variable_types": metadata.readstat_variable_types, "variable_alignment": metadata.variable_alignment, "variable_storage_width": metadata.variable_storage_width, "variable_display_width": metadata.variable_display_width, "variable_measure": metadata.variable_measure } return json.dumps(metadata_json) — Reply to this email directly, view it on GitHub <#228 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/APE2KVTPM2GNTQRH4TEYRE3XFSRQZANCNFSM6AAAAAAX5KDFCI> . You are receiving this because you authored the thread.Message ID: ***@***.***>