added default options to tile_layer

[ozak]

This addresses #123

[birdage]

@ozak can you share an example image or notebook showing the fix pls. 👍

[ozak]

Will do once I solve come issues with the projection of the data I have and upload the web site I am working on.

[birdage]

@ozak great thanks!

[ozak]

One question...I have been trying to have the layer be active to start with, which is what the active=True option should accomplish. Looking at some examples it seemed to be just a mater of having a .addTo(map) in the L.tileLayer() command. But that is not the case from some test I just ran. Any ideas?

[birdage]

@ozak that should do it in my experience, adding add to map should activate the layer auto-magically, can u see the layer in the list?

[ozak]

without the change in the active option that I included, the layer is on the list and can be added by clicking. But if I activate it by adding the .addTo(map) like below does not work.

var tile = L.tileLayer('./data/{z}/{x}/{y}.png',{
                  'minZoom': 0,
                  'maxZoom': 18,
                  'tms': true,
                  'continuousWorld': false,
                  'noWrap': false,
                  'zoomOffset': 0,
                  'zoomReverse': false,
                  'opacity': 1 
                  }).addTo(map);

which is what would be generated with the change I made. Or do you mean another folium command?

My test basically executed the following code in an IPython notebook,

map_osm = folium.Map(location=[0,0], tiles='OpenStreetMap', zoom_start=2)
map_osm.add_tile_layer(tile_name='Name',tile_url='./data/{z}/{x}/{y}.png', active=True, tms=True)
map_osm.add_layers_to_map()
map_osm.create_map(path='./webpage.html')
map_osm.render_iframe = True
map_osm

using master the layer is selectable but not active, and using this pull request it doesn't even work, unless active=False.

[birdage]

@ozak interesting, that would work, maybe its the variable declaration? i.e var tile = try removing that....not hopeful that will work tho

[ozak]

Are you able to have some (non-base) layer active to start with?

[birdage]

@ozak not at the min i dont think, the only things active from init are the base layers

[ozak]

It looks like one needs to add it to the map's layers:

var map = L.map('folium_d128601ca2af4b14b77d02ab215b86b9', {
                             center:[0, 0],
                             zoom: 2,
                             maxBounds: bounds,
                             layers: [base_tile, tilelayer]
                           });

[birdage]

@ozak you shouldn't really need to do that, doing addtomap should add and activate the layer, also if we go the above way it should be part of a layer control group

[ozak]

I completely agree with you. It should add it without having to do much else. Looking at the html output and the code, it seems add_layers_to_map() does add the control, but it is not clear if it also adds the layer. from here I thought it would need another addLayer command. Am I misunderstanding Leaflet?

[ozak]

Looks like one would need to generate some loop to write the elements of layer_list into the following

map.addLayer(layer_layer_element);

which would have to come after the var map= statement.

[ocefpaf]

@ozak Can you add a test to your new functionality? Just a simple template test like we have in https://github.com/python-visualization/folium/blob/master/tests/folium_tests.py would do.

[ozak]

Ok...I added the test and it worked on my computer. I see it is failing on 3.3 (passes on 2.7 and 3.4), but it is not related to what I have done. It fails due to

assert self.map.template_vars['fit_bounds'] == fit_bounds_rendered

[ocefpaf]

Python 3.x is crashing randomly and I am unsure why. Anyway, I re-started the test and Travis-CI is happy now.

Back to you @birdage! If you think this is ready merge away!

[andrewgiessel]

My guess is that the crashes are related to an issue I unearthed in #152: dictionaries are unordered and thus if you don't use the sort_keys=True flag, then the string you get from json.dumps() has random order and the string comparison fails.

[ocefpaf]

@andrewgiessel Thanks for fixing the tests.

@ozak Sorry for the delay! Can you squash your commits?

@birdage Are you done with the reviews?

[ozak]

@ocefpaf Done!

[ocefpaf]

attribution should be a string about the source of the layer, correct? Leave the default equal to None.

[ocefpaf]

@ozak can you write some docstring regarding these options?

[ocefpaf]

Thanks @ozak! I made a few comments in your PR. Feel free to ping me if you need any clarification.

[ozak]

@ocefpaf I'll try to answer to your comments, but won't be able to do so for a couple of days, since I am trying to meet a couple of deadlines. Sorry for the delay!

[ocefpaf]

No problem. The time we demand from contributors is inverse to pay. That means you get zero pay but infinite time 😜

PS: If you address the comments I might get this merged as is and we can work on it later.

[ozak]

The problem is that I have to reacquaint myself with the code and leaflet in general... 😄

[ocefpaf]

The problem is that I have to reacquaint myself with the code and leaflet in general...

I hear you! The whole purpose of folium is to avoid that, but as developer we cannot afford to do it 😒

[ocefpaf]

@ozak with the recent changes on folium I am certain that this belongs in the v0.1.6 branch. Let me know if you can point this PR there or if I should try some git-fu to cherry pick this.

(See http://stackoverflow.com/questions/24159036/how-to-modify-a-pull-request-on-github-to-change-target-branch-to-merge-into)

[ozak]

Sure...I will take a look into it. But as I said it might take a while (at least a month) before I can go back to this.

[ocefpaf]

Not a problem. I am behind the schedule with v0.1.6 (and we can always do a v0.1.7) 😉

[ocefpaf]

Closing this in favor of #236.

Thanks @ozak! (I kept your commits in #236)