pycraf.pathprof ^^^^^^^^^^^^^^ - Add an option to pycraf.pathprof.height_map_data, which allows to retrieve longitude and latitude profiles to each pixel in a map. This could, for example, be used for adding clutter heights along the paths. [#de42379] - Add an option to return raw data from pathprof.gis.regrid_from_geotiff. (May be needed for cases where proper normalisation is required.) [#64]
- Migrate to Cython v3. [#63]
- Use new package structure (pyproject.toml-only). [#65]
pycraf.pathprof ^^^^^^^^^^^^^^ - Add building entry loss model (Rec. ITU-R P.2109) [#d9fd58] - Allow numpy broadcasting for clutter-correction function [#a2984e] and improve its speed via OpenMP [#414233] - Add option to P.452 functions to change base water density of atmosphere [#57, #633c8c]
- Improve the "conversions" Jupyter notebook [#76ae97]
The new pycraf version 2 is mainly about a new packaging structure (using pyproject.toml etc.). Furthermore, the graphical user interface pycraf-gui was moved into a dedicated package. Not every users needs the GUI, but it adds heavily to the package dependencies.
- Add extended IMT composite array antenna (5G) model [#b31fa7]
- Use new pyproj.Transformer class in geospatial sub-package. [#953427]
- Use functools.wraps instead of the astropy wraps. This fixes pycraf for use with astropy version 5. [#683a06]
- Fix matplotlib error with terrain_cmap [#71e284b]
- Add average sectoral antenna pattern for fixed links according to Rec. ITU-R F.1336-5. (The peak sectoral antenna patter was already included, previously.) [#7094f34]
- Add more of the 3GPP TR38.901 path propagation models for IMT (i.e., Urban/ Rural Macro). This is needed for applying power control algorithm related to user equipment in 5G compatibility studies. [#125b25b]
- Fix some bugs related to newer matplotlib version in various tutorial notebooks [#b55c014]
- Add a cache option to pathprof.height_map_data, which is based on the joblib Python package. With this, one can easily re-use existing height data (hprof_cache) from a previous run to save computing time. [#b86c7d]
- Add a generic_heights option to the several functions and the PathProp class in the pathprof sub-package. With this, one case set terrain heights to zero more conveniently. [#a3e799]
- As of Spring 2021, NASA decided to put all SRTM data products behind a log-in page, such that automatic download ceases to work. For the time being, the default server (and only server) in pycraf will thus be viewpano. If you prefer to use NASA tiles (over viewpano), please use their services, e.g., the Land Processes Distributed Active Archive Center.
- In some cases, but only when working with hgt tiles having a different angular resolution than 3", terrain height interpolation was still based on the 3" instead of the proper resolution. [#89e0b0]
- There was a bug in the geometry sub-package, which caused all rotation matrices be transposed (which is the same a applying a negative rotation angle). We used the opportunity to also fix the sign of the etilt angle in the imt2020_composite_pattern. [#47e772,#9fb8d6]
- Fix (stable) HTML manual
- Make satellite sub-package work with sgp4 2.0+
- Add some helper functions to ease working with other GIS data sets. An example would be Landcover data (e.g., from Copernicus mission Corine data set). There is also a tutorial notebook, demonstrating the use. [#29]
- Several functions in the antenna and geometry sub-packages were implemented in Cython and parallelized using OpenMP. This will greatly speed-up Monte-Carlo simulations (e.g., EPFD, which is planned for the next release), where lots of antenna-gain values and boresight angles need to be computed. [#22]
- antenna.imt2020_composite_pattern was using float32 precision, only, for the real and imaginary part (in the complex valued internal calculation). This has been fixed and now uses the complex128 data type. [#26]
- antenna.imt was using wrong units (dimension-less instead of dB) for front-to-back ratios. Note, that all examples still have the same result, as the input parameters were also chosen wrongly. [#25]
- A more detailed error message is now raised, if astropy.units is already imported - which is not allowed with astropy v4 anymore. Please see Issue [#24]
- Migrate CI/CD to Azure Pipelines. This also allows to provide (manylinux-based) binary wheels for Linux on PyPI. [#19]
- A simple graphical user interface was added, which can be used to quickly analyze the path propogation losses along a transmitter-receiver sight-line and plot the path geometry. It is also possible to compute attenuation maps. At the moment, it doesn't come with an amazing amount of features, but it is foreseen to add more in the future. [#15]
- Complete overhaul of the pycraf.atm sub-package [#11]. Now
- it does ray-tracing through atmosphere correctly,
- the computational speed for Annex 1 algorithms (such as ray-tracing) is greatly improved, as performance critical sections are implemented in Cython and physical properties of the atmospheric layers are cached
- there are new functions, such as pycraf.atm.find_elevation, which finds optimal ray geometry (to reach a given target)
- The pycraf.pathprof sub-package was re-worked to improve parallel computations in various places. Most of these are "under the hood", but one new function is losses_complete. It can be used to do bulk calculations of path propagation loss for fixed Tx-Rx location. This makes computations of statistics much faster (see user manual). [#16]
- For pycraf.pathprof we used the BFSG loss (see Rec. ITU-R P.452-16) as proxy for line-of-sight loss. However, to bring it inline with the other quantities returned by the pycraf.pathprof.loss_complete, we now use L_b0p, which includes focussing effects [#13, #16]
- pycraf.geospatial produced an error with newer version of pyproj. Basically, the "+init=" part of the projection definition is now deprecated. We removed this from the pycraf.geospatial module, but as a consequence, pyproj>=2.0 is now required. [#10]
- Make pycraf compatible with astropy>=4.0, which doesn't offer some utils.compat functions any more. [00556d]
- pycraf.protection.ra769_limits now returns an astropy.table.QTable instead of a astropy.table.Table. This ensures that in all circumstances one retrieves proper astropy.units.Quantity objects from the table. Previously, logarithmic units would not fully support this (although this was a just bug in astropy, which is now fixed). [#8]
- Add k-factor to single element IMT2020 pattern in pycraf.antenna module. [#ef1c]
- Add antenna pattern function for IMT advance (LTE) basestation (sectorized, peak-side lobe, see ITU-R Rec. F.1336) to pycraf.antenna module. [#a4e1]
- Various smaller improvements and bugfixes to SRTM querying. [#dcc5, #1950, #0c55]
- The pycraf.geospatial Gauss-Kruger test function revealed a problem with inconsistent results between new proj4 version (5.2.0) and older versions. At the moment, it is not clear, what's going on. The test cases have marked "xfail" for now. [#24e9]
- A sign was wrong in pycraf.antenna IMT2020 composite pattern. [#4164]
- Add a function protection.ra769_calculate_entry that allows to calculate RA.769 thresholds for non-standard values (e.g., to query limits for RAS bands that are not included in the RA.769 tables).
- Update to newest version (v3.0) of Astropy helpers.
- Fixing a serious bug in the B03 tutorial notebook (reflections at wind turbines). Previously, we accounted twice for the distance when calculating the reflected power. As a result, the total received power at the RT is now much larger (though still smaller over direct path).
- Add further utility routines to pycraf.conversions module, to compute antenna temperatures and sensitivities. [#8419]
- Add two new functions atm.elevation_from_airmass and atm.airmass_from_elevation, which use a better formula for small elevations (compared to the 1/sin(El) behavior). Furthermore, the elevation parameter in atm.opacity_from_atten and atm.atten_from_opacity has been made optional. If given, the airmass is corrected for (i.e., one works with zenith opacities). [#61e0]
- The functions pathprof.atten_path_fast and pathprof.atten_map_fast now return dictionaries. This makes it easier to add new return values in the future (without API breaking). Three new parameters are returned for now: the type of path (Line-of-sight or Trans-horizon) and the distance of the horizon w.r.t. Tx/Rx.
- The solution to last exercise in the conversions tutorial notebook was wrong. (Thanks to A. Jessner for spotting this.)
- The phi = 0-singularity if using do_bessel=True in antenna.ras_pattern was not properly handled. [#5acf]
- Added some notebooks with exercises and solutions, as well as tutorial 03e.
- Add correlation-level parameter to imt2020_composite_pattern function. [#c57b]
- Add various convenience functions to create 3D rotation matrices from rotation axis or Euler angles (and vice versa). Streamline the geometry subpackage to allow proper numpy broadcasting. [#c970]
- Add a method, geoid_area to calculate surface area on the WGS84 ellipsoid. Only rectangular limits (absolute coordinates) are supported. This can be used to determine the area of SRTM pixels (in km^2). [#678f]
- Add a method atten_path_fast, which can be used to calculate attenuations for full paths very quickly (compared to the manual approach); also see tutorial notebooks. To produce the necessary aux-data dictionary, two functions are available: height_path_data and height_path_data_generic. To avoid confusion and to streamline everything, the function heigth_profile_data was renamed to height_map_data. [#9d6a]
pycraf.protection ^^^^^^^^^^^^^^^ - Add the possibility to generate VLBI threshold values in ra769_limits function, as contained in Table 3 of RA.769. Furthermore, it is now possible to specify the integration time to be used for the thresholds. [#8a15]
- Add a notebook about how to tilt or rotate IMT2020 antenna patterns.
- Various updates.
- In atm.atten_slant_annex1 the obs_alt parameter was not properly accounted for. This led to significant errors for high-altitude observers. [#f616]
- The function imt2020_composite_pattern in the antenna subpackage now allows better broadcasting of input arrays. Speed was also improved. [#b8ac, #1219]
- This sub-package was heavily re-factored. One can now work with EPSG or ESRI codes (see docs) and there is a factory to produce arbitrary transforms, which come with correct docstrings and quantity/range checker (i.e, proper unit handling). Also, we finally made ITRF work. [#6892]
- Increase minimal numpy version to 1.11 to avoid build conflicts with MacOS and Windows wheels. The wheels are now built with this minimal numpy version (1.11) rather than the latest version. Users of wheels have to have at least the same numpy version as the one with which pycraf wheels were built. [#2139]
- Add option interp to SrtmConf to allow different interpolation schemes for SRTM data. Currently, 'nearest', 'linear' (default), and 'spline' are supported. [#8b43]
- SRTM query functions now support numpy broadcasting. [#af45]
- SRTM-related plots in documentation were not rendered. [#ea01]
- Clutter loss for "CLUTTER.UNKNOWN" was buggy (must always be zero). [#a8b5]
- Tests now don't run any SRTM-data related function, only if the option remote-data=any is given (on CLI, within Python it is invoked with pycraf.test(remote_data='any')). Therefore, one doesn't need to manually download SRTM tiles beforehand anymore. [#eeab]
- Minimize dependencies: for a some of pycraf sub-packages, do imports only on demand and not during pycraf importing. Examples are the satellite and geospatial sub-packages, that depend on pyproj and sgp4. Of course, to use these sub-packages, one has to install the necessary dependencies. [#8051]
Tests that need to download data from the internet are now decorated with astropy's @remote_data decorator. If you want to run these tests, use:
import pycraf pycraf.test(remote_data='any')or
python setup.py test --remote-data='any'
otherwise, these will be skipped over. [#1444]
- MacOS tests finally work. [#239d]
- Much better handling of SRTM data was implemented. It is now possible to define the SRTM directory during run-time. Furthermore, one can have pycraf download missing tiles. For this a new SrtmConf manager was introduced. [#2d30, #01ba, #208c, #2b5d]