LaTeX Generation


One of the features the TOM Toolkit offers is automated generation of LaTeX-formatted data tables. The LaTeX table tool allows the user to select the parameters for an entity in their TOM–for example, a Target–and generate a table of those parameters for all targets within a list. At the moment, the Toolkit supports table generation for two built-in models–ObservationGroups and TargetLists.

A LaTeX processor can be created for any model, or, with some additional modifications, any combination of models. The supported LaTeX processors must be specified in settings.py in the TOM_LATEX_PROCESSORS as key/value pairs, with the model being the key, and the processor class being the value. By default, the following processors are automatically present in settings.py:

TOM_LATEX_PROCESSORS = {
    'ObservationGroup': 'tom_publications.processors.latex_processor.ObservationGroupLatexProcessor',
    'TargetList': 'tom_publications.processors.target_list_latex_processor.TargetListLatexProcessor'
}

Custom Processing


The built-in LaTeX table generation is good, but it certainly has some shortcomings, and can’t be expected to cover every or even most use cases. As such, the implementation allows for smooth addition of any custom processing.

In order to generate a LaTeX table for a unique use case, we’ll need to write a custom LaTeX processor, which we’ll go through below. A LaTeX processor has a custom Form class and a Processor class, and the Processor class has a function which takes data from your TOM DB and outputs it in the preferred LaTeX-formatted table. To begin, here’s a brief look at part of the structure of the tom_publications app in the TOM Toolkit:

tom_publications
├──latex.py
└──processors
   ├──target_list_latex_processor.py
   └──observation_group_latex_processor.py

Perhaps one wants a processor that generates a table simply for all the photometric or spectroscopic data for a given target. The first thing to be done is to create a target_photometry_latex_processor.py. We’ll create a new file for our processor, and then create a TargetListLatexProcessor class that inherits from GenericLatexProcessor. GenericLatexProcessor has an abstract method that must be implemented called create_latex, so we’ll also add that:

from tom_publications.latex import GenericLatexProcessor

class TargetDataLatexProcessor(GenericLatexProcessor):

  def create_latex(self, cleaned_data):
    pass

The GenericLatexProcessor also has a form class that renders the correct set of fields to be generated. In our case, we’d like the user to be able to choose between spectroscopy or photometry. So let’s create the form. We’ll also create one form field, and populate it with our two choices:

from django import forms

from tom_publications.latex import GenericLatexProcessor, GenericLatexForm


class TargetDataLatexForm(GenericLatexForm):
  data_type = forms.ChoiceField(
    choices=[('spectroscopy', 'Spectroscopy'), ('photometry', 'Photometry')],
    required=True,
    widget=forms.RadioSelect()
  )


class TargetDataLatexProcessor(GenericLatexProcessor):
...

With the form implemented, we can implement our create_latex method and add our TargetDataLatexForm as the form_class.

The base form class always includes model_pk, which gives us a way to access the object for which we’re generating data.

import json

from django import forms

from tom_dataproducts.models import ReducedDatum
from tom_publications.latex import GenericLatexProcessor, GenericLatexForm
from tom_targets.models import Target

...

class TargetDataLatexProcessor(GenericLatexProcessor):
  form_class = TargetDataLatexForm

  def create_latex_table_data(self, cleaned_data):
    target = Target.objects.get(pk=cleaned_data.get('model_pk'))
    data = ReducedDatum.objects.filter(target=target, data_type=cleaned_data.get('data_type'))

    table_data = {}
    if cleaned_data.get('data_type') == 'photometry':
      for datum in data:
        for key, value in json.loads(datum.value).items():
          table_data.setdefault(key, []).append(value)
    elif cleaned_data.get('data_type') == 'spectroscopy':
      ...

    return table_data

The above example only shows the photometric table generation, but spectroscopic can be left as an exercise to the reader.

The last two steps are to link our new processor to our existing code. First, in our settings.py (making sure you replace the displayed path with the correct one for your TOM):

...
TOM_LATEX_PROCESSORS = {
    'ObservationGroup': 'tom_publications.processors.latex_processor.ObservationGroupLatexProcessor',
    'TargetList': 'tom_publications.processors.target_list_latex_processor.TargetListLatexProcessor',
    'Target': 'tom_publications.processors.target_data_latex_processor.TargetDataLatexProcessor'
}
...

We add a Target processor. For the default implementation, all processors must be tied to a TOM model, but with a custom templatetag (or enough requests to the developers), it can be expanded further.

Then, in our overridden target_detail.html template (details on overriding templates can be found here), we add a button:

...
<div id="target-info">
      {% target_feature object %}
      {% latex_button object %}
      {% if object.future_observations %}
...

For context, the template tag being referenced by {% latex_button object %} can be seen below. It accepts an instance of a model from your TOM and generates a button with the correct query parameters to send to your form.

@register.inclusion_tag('tom_publications/partials/latex_button.html')
def latex_button(object):
    """
    Renders a button that redirects to the LaTeX table generation page for the specified model instance. Requires an
    object, which is generally the object in the context for the page on which the templatetag will be used.
    """
    model_name = object._meta.label
    return {'model_name': object._meta.label, 'model_pk': object.id}

With all that done, you will now be able to generate tables of photometric (and eventually spectroscopic) data of any target in your TOM. Here’s our final target_data_latex_processor.py:

import json

from django import forms

from tom_dataproducts.models import ReducedDatum
from tom_publications.latex import GenericLatexProcessor, GenericLatexForm
from tom_targets.models import Target


class TargetDataLatexForm(GenericLatexForm):
    data_type = forms.ChoiceField(
        choices=[('spectroscopy', 'Spectroscopy'), ('photometry', 'Photometry')],
        required=True,
        widget=forms.RadioSelect()
    )


class TargetDataLatexProcessor(GenericLatexProcessor):
    form_class = TargetDataLatexForm

    def create_latex_table_data(self, cleaned_data):
        target = Target.objects.get(pk=cleaned_data.get('model_pk'))
        data = ReducedDatum.objects.filter(target=target, data_type=cleaned_data.get('data_type'))

        table_data = {}
        if cleaned_data.get('data_type') == 'photometry':
            for datum in data:
                for key, value in json.loads(datum.value).items():
                    table_data.setdefault(key, []).append(value)
        elif cleaned_data.get('data_type') == 'spectroscopy':
            ...

        return table_data