I am trying profiling my code in GPflow 2 as I need to know which part of my code consumes the most CPU time. In GPflow 1 there was a gpflowrc file where you could set dump_timeline = True but this has changed in GPflow 2 (to the gpflow.config module) and I can't find a similar option in there.
Working with TensorFlow 2 is much simpler than TensorFlow 1, so with GPflow 2 we're relying a lot more on TensorFlow built-ins instead of adding extra code - GPflow 2 is "just another TensorFlow graph". So you should be able to directly use the TensorFlow profiler: see this blog post for an introduction and the guide in the TensorFlow documentation for more details.
(According to https://github.com/tensorflow/tensorboard/issues/2874, TensorBoard's "Trace Viewer" for the timeline should now work fine in Firefox, but if you encounter any issues on the visualization side it's worth trying out Chrome.)
Related
I am training a Neural Network model with pytorch.
Since this model is very complicated, I made use of the pytorch_xla package to use TPU. I finished training the model and now I want to save the weight so I will be able to use them from any envrionment.
I tried to save the data like so
file_name = "model_params"
torch.save(model.state_dict(), file_name)
and when I tried to load them (from environment which does not support TPU)
model.load_state_dict(torch.load(file name))
I got the following error
NotImplementedError: Could not run 'aten::empty_strided' with arguments from the 'XLA' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'aten::empty_strided' is only available for these backends: [CPU, Meta, BackendSelect, Named, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradXLA, UNKNOWN_TENSOR_TYPE_ID, AutogradMLC, AutogradHPU, AutogradNestedTensor, AutogradPrivateUse1, AutogradPrivateUse2, AutogradPrivateUse3, Tracer, Autocast, Batched, VmapMode].
Is there a way to do what I want?
I use Python for plotting geospatial data on maps.
For certain map-styles, such as ["basic", "streets", "outdoors", "light", "dark", "satellite", "satellite-streets"], I need a mapbox-access token and for some geospatial plotting packages like folium I even need to create my own link for retrieving the map-tiles.
So far, it worked great with the style "satellite":
mapbox_style = "satellite"
mapbox_access_token = "....blabla"
request_link = f"https://api.mapbox.com/v4/mapbox.{mapbox_style}/{{z}}/{{x}}/{{y}}#2x.jpg90?access_token={mapbox_access_token}"
However, when choosing "satellite-streets" as mapbox-tile-ID, the output doesn't show a background map anymore. It fails with inserting "satellite-streets", "satellitestreets" and "satellite_streets" into the aforementioned link-string.
Why is that and how can I come to know what's the correct tile-ID-name for "satellite-streets"?
I found an answer when reaching out to the customer support.
Apparently, one has to access the static APIs which have specific names listed on their website:
"In general, the styles that you mentioned including
"satellite_streets" that you are referencing are our classic styles
that are going to be deprecated starting June 1st. I would recommend
using our modern static API the equivalent modern styles. This
will allow you to see the most updated street data as well.
Like the example request below:
https://api.mapbox.com/styles/v1/mapbox/satellite-streets-v11/tiles/1/1/0?access_token={your_token}
Here is more info on the deprecation of the classic styles and
the migration guide for them."
My personal adaptation after having tried everything out myself, is:
Via combining the above-mentioned with the details on how to construct a Mapbox-request link on this documention from mapbox' website,
I finally managed to make it work.
An example request looks like so (in python using f-strings):
mapbox_tile_URL = f"https://api.mapbox.com/styles/v1/mapbox/{tileset_ID_str}/tiles/{tilesize_pixels}/{{z}}/{{x}}/{{y}}#2x?access_token={mapbox_access_token}"
The tileset_ID_str could be e.g. "satellite-streets-v11" which can be seen at the following link containing valid static maps.
I hava a quick question:
I see that you have the high-level charts in the version 0.11.0:
http://docs.bokeh.org/en/0.11.0/docs/user_guide/charts.html
But I can't find the same topic in the last version (1.0.3)? Did bokeh team remove it?
I have been working on adapting histograms to my project but can't find the histograms section in the last version?
Any ideas?
The bokeh.charts API was deprecated and scheduled to be removed in version 1.0 several years ago. However, even that schedule was also eventually accelerated, due to lack of interest and better alternatives. It has been completely gone from the project since late 2017.
For very high level APIs on top of Bokeh, consider Chartify or Holoviews. Or just create histograms directly with the stable bokeh.plotting API. e.g.
from bokeh.plotting import figure, output_file, show
from bokeh.sampledata.autompg import autompg as df
from numpy import histogram
p = figure(plot_height=300)
hist, edges = histogram(df.hp, density=True, bins=20)
p.quad(top=hist, bottom=0, left=edges[:-1], right=edges[1:], line_color="white")
show(p)
I have a 3-axis system, which resembles the CNC-axis, and a PowerPanel from B&R. Now I plan to execute the g-code, which I read from a file, which is fe located in the Downloads folder. This will be done with fbs from B&R Automation Studio.
I got confused with the help, since it says in drive engineering/mapp Motion/technologie/function block/mpcnc3axis/description the part of moveProgram, where it redirects you to MC_BR_MoveProgram \ description. There stands that you can start the program with [Program name prefix] [File device name:] [Path] Program name syntax. Now I am not sure how to include this with my MpCnc3Axis code or if I can use the syntax also in MpCnc3AxisParType ProgramName part to start the program from fe Downloads Folder, since it didn't work with McPrgAdvParType InitProgramName.
Any links to related material would be appreciated.
I would also accept a solution with C, if it is easier.
Depending on if you are willing to buy software licences from B&R, running a CNC system is more or less easy.
The state of the art implementation is with mappMotion. It requires software licenses and gives you ready to go FUBs to execute nc files, jogging, direct movements, etc. This works the same way for different robot types and CNC kinematics. I recently uploaded a small implementation of a robot on GitLab: https://gitlab.com/kirni/bur_robotic_sample
The alternative is the old ARNC0 approach. No more new features are implemented, but it is free - as long as you do not need any functions that require a Dual Use agreement (i.e. 5 Axis CNC, etc.). The implementation is rather complex and is definitely to extensive to be answered completely at stackoverflow.
However Automation Studio ships with a sample for a two axis CNC. In the logical view you can add a new object Library Samples which opens a dialog where you can select Samples/Motion/CNC system (2 axis)/LibARNC0_Basic_ST.zip. This would be the best starting point, which is also documented in the help in MotionControl/ACP10 ARNC0/Examples/CNC or guid: 4620bb1c-8694-4f27-87d0-805686d7c0c8 as of AS 4.4.4.64.
I hope this helps!
I want to make a network graph which shows the distribution of our documents in our folder structure.
I have the nodefile, edgefile and gephi graph file in this location:
https://1drv.ms/f/s!AuVfRBdVHkO7hgs5K9r9f7jBBAUH
What I do is:
Run the algorithm ForceAtlas2 with scaling 10-20, dissuade hub marked and prevent overlap marked, all other standard setting.
What I get is a graph with groups radial/spherical distributed. However, what I want is a tree directed network graph.
Anyone know how I can adjust Gephi to make this?
Thanks!
I just found a solution.
I tested the file format as shown on the Yed site "import excel file" page
http://yed.yworks.com/support/manual/import_excel.html
This gave me the Yed import dialog (took a life time to figure out that it's a pop up menu and not selectable through the standard menu)
Anyway, it worked and I've adjusted the test files with the data prepared for the Gehpi. This was pretty easy, I could used the source target ID's etc. Just copy paste.
I load it into Yed and used some directed and radial clustering algorithms on it. Works fine!
Below you can find the excel node/edge file used to import in Yed and the graph file you can open with Yed to see the final radial result.
https://1drv.ms/f/s!AuVfRBdVHkO7hg6DExK_eVkm5_mR
Only thing to figure out is how to combine the weight (which represents the number of documents) with the node size.
Unfortunately, as of version 0.9.0, Gephi no longer supports hierarchical graphs. Maybe try using a previous version?
Other alternatives involve more complex software, such as Graphviz, but you need a .dot file instead of your .csv. I looked all over, but could not find an easy-to-use csv to dot converter.
You could try looking at d3-hierarchy, a node.js program, but then again you need to use the not-so-user-friendly npm. If you look at the link, it looks like it can produce the kind of diagram you're looking for.