healpy tutorial¶
See the Jupyter Notebook version of this tutorial at https://github.com/healpy/healpy/blob/master/doc/tutorial.ipynb
Choose the inline
backend of maptlotlib
to display the plots inside the Jupyter Notebook
[1]:
import matplotlib.pyplot as plt
%matplotlib inline
[2]:
import numpy as np
import healpy as hp
NSIDE and ordering¶
Maps are simply numpy arrays, where each array element refers to a location in the sky as defined by the Healpix pixelization schemes (see the healpix website).
Note: Running the code below in a regular Python session will not display the maps; it’s recommended to use an IPython shell or a Jupyter notebook.
The resolution of the map is defined by the NSIDE parameter, which is generally a power of 2.
[3]:
NSIDE = 32
print(
"Approximate resolution at NSIDE {} is {:.2} deg".format(
NSIDE, hp.nside2resol(NSIDE, arcmin=True) / 60
)
)
Approximate resolution at NSIDE 32 is 1.8 deg
The function healpy.pixelfunc.nside2npix
gives the number of pixels NPIX of the map:
[4]:
NPIX = hp.nside2npix(NSIDE)
print(NPIX)
12288
The same pixels in the map can be ordered in 2 ways, either RING, where they are numbered in the array in horizontal rings starting from the North pole:
[5]:
m = np.arange(NPIX)
hp.mollview(m, title="Mollview image RING")
hp.graticule()
The standard coordinates are the colatitude \(\theta\), \(0\) at the North Pole, \(\pi/2\) at the equator and \(\pi\) at the South Pole and the longitude \(\phi\) between \(0\) and \(2\pi\) eastward, in a Mollview projection, \(\phi=0\) is at the center and increases eastward toward the left of the map.
We can also use vectors to represent coordinates, for example vec
is the normalized vector that points to \(\theta=\pi/2, \phi=3/4\pi\):
[6]:
vec = hp.ang2vec(np.pi / 2, np.pi * 3 / 4)
print(vec)
[-7.07106781e-01 7.07106781e-01 6.12323400e-17]
We can find the indices of all the pixels within \(10\) degrees of that point and then change the value of the map at those indices:
[7]:
ipix_disc = hp.query_disc(nside=32, vec=vec, radius=np.radians(10))
[8]:
m = np.arange(NPIX)
m[ipix_disc] = m.max()
hp.mollview(m, title="Mollview image RING")
We can retrieve colatitude and longitude of each pixel using pix2ang
, in this case we notice that the first 4 pixels cover the North Pole with pixel centers just ~\(1.5\) degrees South of the Pole all at the same latitude. The fifth pixel is already part of another ring of pixels.
[9]:
theta, phi = np.degrees(hp.pix2ang(nside=32, ipix=[0, 1, 2, 3, 4]))
[10]:
theta
[10]:
array([1.46197116, 1.46197116, 1.46197116, 1.46197116, 2.92418036])
[11]:
phi
[11]:
array([ 45. , 135. , 225. , 315. , 22.5])
The RING ordering is necessary for the Spherical Harmonics transforms, the other option is NESTED ordering which is very efficient for map domain operations because scaling up and down maps is achieved just multiplying and rounding pixel indices. See below how pixel are ordered in the NESTED scheme, notice the structure of the 12 HEALPix base pixels (NSIDE 1):
[12]:
m = np.arange(NPIX)
hp.mollview(m, nest=True, title="Mollview image NESTED")
All healpy
routines assume RING ordering, in fact as soon as you read a map with read_map
, even if it was stored as NESTED, it is transformed to RING. However, you can work in NESTED ordering passing the nest=True
argument to most healpy
routines.
Reading and writing maps to file¶
For the following section, it is required to download larger maps by executing from the terminal the bash script healpy_get_wmap_maps.sh
which should be available in your path.
This will download the higher resolution WMAP data into the current directory.
[13]:
!wget -c http://lambda.gsfc.nasa.gov/data/map/dr4/skymaps/7yr/raw/wmap_band_iqumap_r9_7yr_W_v4.fits;wget -c http://lambda.gsfc.nasa.gov/data/map/dr4/ancillary/masks/wmap_temperature_analysis_mask_r9_7yr_v4.fits
--2024-03-19 15:41:35-- http://lambda.gsfc.nasa.gov/data/map/dr4/skymaps/7yr/raw/wmap_band_iqumap_r9_7yr_W_v4.fits
Resolving lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)...
129.164.179.68, 2001:4d0:2310:150::68
Connecting to lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)|129.164.179.68|:80... connected.
HTTP request sent, awaiting response...
301 Moved Permanently
Location: https://lambda.gsfc.nasa.gov/data/map/dr4/skymaps/7yr/raw/wmap_band_iqumap_r9_7yr_W_v4.fits [following]
--2024-03-19 15:41:35-- https://lambda.gsfc.nasa.gov/data/map/dr4/skymaps/7yr/raw/wmap_band_iqumap_r9_7yr_W_v4.fits
Connecting to lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)|129.164.179.68|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 100676160 (96M)
Saving to: ‘wmap_band_iqumap_r9_7yr_W_v4.fits’
wmap_band 0%[ ] 0 --.-KB/s
wmap_band_ 1%[ ] 1.66M 8.13MB/s
wmap_band_i 3%[ ] 3.67M 8.93MB/s
wmap_band_iq 6%[> ] 5.78M 9.41MB/s
wmap_band_iqu 8%[> ] 7.89M 9.64MB/s
wmap_band_iqum 10%[=> ] 9.89M 9.67MB/s
- wmap_band_iquma 12%[=> ] 11.99M 9.78MB/s
</pre>
- wmap_band_iquma 12%[=> ] 11.99M 9.78MB/s
end{sphinxVerbatim}
wmap_band_iquma 12%[=> ] 11.99M 9.78MB/s
- more-to-come:
- wmap_band_iqumap 14%[=> ] 14.10M 9.86MB/s
</pre>
- wmap_band_iqumap 14%[=> ] 14.10M 9.86MB/s
end{sphinxVerbatim}
wmap_band_iqumap 14%[=> ] 14.10M 9.86MB/s
- more-to-come:
- wmap_band_iqumap_ 16%[==> ] 15.97M 9.77MB/s
</pre>
- wmap_band_iqumap_ 16%[==> ] 15.97M 9.77MB/s
end{sphinxVerbatim}
wmap_band_iqumap_ 16%[==> ] 15.97M 9.77MB/s
- more-to-come:
- wmap_band_iqumap_r 18%[==> ] 18.07M 9.83MB/s
</pre>
- wmap_band_iqumap_r 18%[==> ] 18.07M 9.83MB/s
end{sphinxVerbatim}
wmap_band_iqumap_r 18%[==> ] 18.07M 9.83MB/s
- wmap_band_iqumap_r9 21%[===> ] 20.16M 9.88MB/s
</pre>
- wmap_band_iqumap_r9 21%[===> ] 20.16M 9.88MB/s
end{sphinxVerbatim}
wmap_band_iqumap_r9 21%[===> ] 20.16M 9.88MB/s
- map_band_iqumap_r9_ 23%[===> ] 22.15M 9.88MB/s
</pre>
- map_band_iqumap_r9_ 23%[===> ] 22.15M 9.88MB/s
end{sphinxVerbatim}
map_band_iqumap_r9_ 23%[===> ] 22.15M 9.88MB/s
- ap_band_iqumap_r9_7 25%[====> ] 24.20M 9.91MB/s
</pre>
- ap_band_iqumap_r9_7 25%[====> ] 24.20M 9.91MB/s
end{sphinxVerbatim}
ap_band_iqumap_r9_7 25%[====> ] 24.20M 9.91MB/s
- p_band_iqumap_r9_7y 27%[====> ] 26.29M 9.95MB/s
</pre>
- p_band_iqumap_r9_7y 27%[====> ] 26.29M 9.95MB/s
end{sphinxVerbatim}
p_band_iqumap_r9_7y 27%[====> ] 26.29M 9.95MB/s
- _band_iqumap_r9_7yr 29%[====> ] 28.33M 9.96MB/s
</pre>
- _band_iqumap_r9_7yr 29%[====> ] 28.33M 9.96MB/s
end{sphinxVerbatim}
_band_iqumap_r9_7yr 29%[====> ] 28.33M 9.96MB/s
- band_iqumap_r9_7yr_ 31%[=====> ] 30.43M 9.99MB/s eta 7s
</pre>
- band_iqumap_r9_7yr_ 31%[=====> ] 30.43M 9.99MB/s eta 7s
end{sphinxVerbatim}
band_iqumap_r9_7yr_ 31%[=====> ] 30.43M 9.99MB/s eta 7s
- and_iqumap_r9_7yr_W 33%[=====> ] 32.54M 10.1MB/s eta 7s
</pre>
- and_iqumap_r9_7yr_W 33%[=====> ] 32.54M 10.1MB/s eta 7s
end{sphinxVerbatim}
and_iqumap_r9_7yr_W 33%[=====> ] 32.54M 10.1MB/s eta 7s
- nd_iqumap_r9_7yr_W_ 35%[======> ] 34.56M 10.1MB/s eta 7s
</pre>
- nd_iqumap_r9_7yr_W_ 35%[======> ] 34.56M 10.1MB/s eta 7s
end{sphinxVerbatim}
nd_iqumap_r9_7yr_W_ 35%[======> ] 34.56M 10.1MB/s eta 7s
- d_iqumap_r9_7yr_W_v 38%[======> ] 36.67M 10.1MB/s eta 7s
</pre>
- d_iqumap_r9_7yr_W_v 38%[======> ] 36.67M 10.1MB/s eta 7s
end{sphinxVerbatim}
d_iqumap_r9_7yr_W_v 38%[======> ] 36.67M 10.1MB/s eta 7s
- _iqumap_r9_7yr_W_v4 40%[=======> ] 38.77M 10.1MB/s eta 7s
</pre>
- _iqumap_r9_7yr_W_v4 40%[=======> ] 38.77M 10.1MB/s eta 7s
end{sphinxVerbatim}
_iqumap_r9_7yr_W_v4 40%[=======> ] 38.77M 10.1MB/s eta 7s
- iqumap_r9_7yr_W_v4. 42%[=======> ] 40.88M 10.2MB/s eta 5s
</pre>
- iqumap_r9_7yr_W_v4. 42%[=======> ] 40.88M 10.2MB/s eta 5s
end{sphinxVerbatim}
iqumap_r9_7yr_W_v4. 42%[=======> ] 40.88M 10.2MB/s eta 5s
- qumap_r9_7yr_W_v4.f 44%[=======> ] 42.97M 10.2MB/s eta 5s
</pre>
- qumap_r9_7yr_W_v4.f 44%[=======> ] 42.97M 10.2MB/s eta 5s
end{sphinxVerbatim}
qumap_r9_7yr_W_v4.f 44%[=======> ] 42.97M 10.2MB/s eta 5s
- umap_r9_7yr_W_v4.fi 46%[========> ] 45.07M 10.2MB/s eta 5s
</pre>
- umap_r9_7yr_W_v4.fi 46%[========> ] 45.07M 10.2MB/s eta 5s
end{sphinxVerbatim}
umap_r9_7yr_W_v4.fi 46%[========> ] 45.07M 10.2MB/s eta 5s
- map_r9_7yr_W_v4.fit 49%[========> ] 47.05M 10.2MB/s eta 5s
</pre>
- map_r9_7yr_W_v4.fit 49%[========> ] 47.05M 10.2MB/s eta 5s
end{sphinxVerbatim}
map_r9_7yr_W_v4.fit 49%[========> ] 47.05M 10.2MB/s eta 5s
- ap_r9_7yr_W_v4.fits 51%[=========> ] 49.17M 10.2MB/s eta 5s
</pre>
- ap_r9_7yr_W_v4.fits 51%[=========> ] 49.17M 10.2MB/s eta 5s
end{sphinxVerbatim}
ap_r9_7yr_W_v4.fits 51%[=========> ] 49.17M 10.2MB/s eta 5s
- p_r9_7yr_W_v4.fits 53%[=========> ] 51.15M 10.2MB/s eta 4s
</pre>
- p_r9_7yr_W_v4.fits 53%[=========> ] 51.15M 10.2MB/s eta 4s
end{sphinxVerbatim}
p_r9_7yr_W_v4.fits 53%[=========> ] 51.15M 10.2MB/s eta 4s
- _r9_7yr_W_v4.fits 55%[==========> ] 53.25M 10.2MB/s eta 4s
</pre>
- _r9_7yr_W_v4.fits 55%[==========> ] 53.25M 10.2MB/s eta 4s
end{sphinxVerbatim}
_r9_7yr_W_v4.fits 55%[==========> ] 53.25M 10.2MB/s eta 4s
- r9_7yr_W_v4.fits 57%[==========> ] 55.23M 10.2MB/s eta 4s
</pre>
- r9_7yr_W_v4.fits 57%[==========> ] 55.23M 10.2MB/s eta 4s
end{sphinxVerbatim}
r9_7yr_W_v4.fits 57%[==========> ] 55.23M 10.2MB/s eta 4s
- 9_7yr_W_v4.fits 59%[==========> ] 57.33M 10.2MB/s eta 4s
</pre>
- 9_7yr_W_v4.fits 59%[==========> ] 57.33M 10.2MB/s eta 4s
end{sphinxVerbatim}
9_7yr_W_v4.fits 59%[==========> ] 57.33M 10.2MB/s eta 4s
- _7yr_W_v4.fits 61%[===========> ] 59.44M 10.2MB/s eta 4s
</pre>
- _7yr_W_v4.fits 61%[===========> ] 59.44M 10.2MB/s eta 4s
end{sphinxVerbatim}
_7yr_W_v4.fits 61%[===========> ] 59.44M 10.2MB/s eta 4s
- 7yr_W_v4.fits 64%[===========> ] 61.50M 10.1MB/s eta 3s
</pre>
- 7yr_W_v4.fits 64%[===========> ] 61.50M 10.1MB/s eta 3s
end{sphinxVerbatim}
7yr_W_v4.fits 64%[===========> ] 61.50M 10.1MB/s eta 3s
- yr_W_v4.fits 66%[============> ] 63.61M 10.2MB/s eta 3s
</pre>
- yr_W_v4.fits 66%[============> ] 63.61M 10.2MB/s eta 3s
end{sphinxVerbatim}
yr_W_v4.fits 66%[============> ] 63.61M 10.2MB/s eta 3s
- r_W_v4.fits 68%[============> ] 65.65M 10.2MB/s eta 3s
</pre>
- r_W_v4.fits 68%[============> ] 65.65M 10.2MB/s eta 3s
end{sphinxVerbatim}
r_W_v4.fits 68%[============> ] 65.65M 10.2MB/s eta 3s
- _W_v4.fits 70%[=============> ] 67.70M 10.2MB/s eta 3s
</pre>
- _W_v4.fits 70%[=============> ] 67.70M 10.2MB/s eta 3s
end{sphinxVerbatim}
_W_v4.fits 70%[=============> ] 67.70M 10.2MB/s eta 3s
- W_v4.fits 72%[=============> ] 69.76M 10.2MB/s eta 3s
</pre>
- W_v4.fits 72%[=============> ] 69.76M 10.2MB/s eta 3s
end{sphinxVerbatim}
W_v4.fits 72%[=============> ] 69.76M 10.2MB/s eta 3s
- _v4.fits 74%[=============> ] 71.85M 10.2MB/s eta 2s
</pre>
- _v4.fits 74%[=============> ] 71.85M 10.2MB/s eta 2s
end{sphinxVerbatim}
_v4.fits 74%[=============> ] 71.85M 10.2MB/s eta 2s
- v4.fits 76%[==============> ] 73.92M 10.2MB/s eta 2s
</pre>
- v4.fits 76%[==============> ] 73.92M 10.2MB/s eta 2s
end{sphinxVerbatim}
v4.fits 76%[==============> ] 73.92M 10.2MB/s eta 2s
- 4.fits 79%[==============> ] 75.98M 10.2MB/s eta 2s
</pre>
- 4.fits 79%[==============> ] 75.98M 10.2MB/s eta 2s
end{sphinxVerbatim}
4.fits 79%[==============> ] 75.98M 10.2MB/s eta 2s
- .fits 81%[===============> ] 78.06M 10.2MB/s eta 2s
</pre>
- .fits 81%[===============> ] 78.06M 10.2MB/s eta 2s
end{sphinxVerbatim}
.fits 81%[===============> ] 78.06M 10.2MB/s eta 2s
- fits 83%[===============> ] 80.17M 10.2MB/s eta 2s
</pre>
- fits 83%[===============> ] 80.17M 10.2MB/s eta 2s
end{sphinxVerbatim}
fits 83%[===============> ] 80.17M 10.2MB/s eta 2s
- its 85%[================> ] 82.27M 10.2MB/s eta 1s
</pre>
- its 85%[================> ] 82.27M 10.2MB/s eta 1s
end{sphinxVerbatim}
its 85%[================> ] 82.27M 10.2MB/s eta 1s
- ts 87%[================> ] 84.37M 10.3MB/s eta 1s
</pre>
- ts 87%[================> ] 84.37M 10.3MB/s eta 1s
end{sphinxVerbatim}
ts 87%[================> ] 84.37M 10.3MB/s eta 1s
- s 90%[=================> ] 86.44M 10.3MB/s eta 1s
</pre>
- s 90%[=================> ] 86.44M 10.3MB/s eta 1s
end{sphinxVerbatim}
s 90%[=================> ] 86.44M 10.3MB/s eta 1s
92%[=================> ] 88.49M 10.3MB/s eta 1s
w 94%[=================> ] 90.46M 10.2MB/s eta 1s
wm 96%[==================> ] 92.57M 10.3MB/s eta 0s
wma 98%[==================> ] 94.68M 10.3MB/s eta 0s
wmap_band_iqumap_r9 100%[===================>] 96.01M 10.3MB/s in 9.5s
2024-03-19 15:41:45 (10.1 MB/s) - ‘wmap_band_iqumap_r9_7yr_W_v4.fits’ saved [100676160/100676160]
URL transformed to HTTPS due to an HSTS policy –2024-03-19 15:41:45– https://lambda.gsfc.nasa.gov/data/map/dr4/ancillary/masks/wmap_temperature_analysis_mask_r9_7yr_v4.fits Resolving lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)… 129.164.179.68, 2001:4d0:2310:150::68 Connecting to lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)|129.164.179.68|:443… connected. </pre>
wmap_band_iqumap_r9 100%[===================>] 96.01M 10.3MB/s in 9.5s
2024-03-19 15:41:45 (10.1 MB/s) - ‘wmap_band_iqumap_r9_7yr_W_v4.fits’ saved [100676160/100676160]
URL transformed to HTTPS due to an HSTS policy –2024-03-19 15:41:45– https://lambda.gsfc.nasa.gov/data/map/dr4/ancillary/masks/wmap_temperature_analysis_mask_r9_7yr_v4.fits Resolving lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov){ldots} 129.164.179.68, 2001:4d0:2310:150::68 Connecting to lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)|129.164.179.68|:443{ldots} connected. end{sphinxVerbatim}
wmap_band_iqumap_r9 100%[===================>] 96.01M 10.3MB/s in 9.5s
2024-03-19 15:41:45 (10.1 MB/s) - ‘wmap_band_iqumap_r9_7yr_W_v4.fits’ saved [100676160/100676160]
URL transformed to HTTPS due to an HSTS policy –2024-03-19 15:41:45– https://lambda.gsfc.nasa.gov/data/map/dr4/ancillary/masks/wmap_temperature_analysis_mask_r9_7yr_v4.fits Resolving lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)… 129.164.179.68, 2001:4d0:2310:150::68 Connecting to lambda.gsfc.nasa.gov (lambda.gsfc.nasa.gov)|129.164.179.68|:443… connected.
HTTP request sent, awaiting response... 200 OK
Length: 25174080 (24M)
Saving to: ‘wmap_temperature_analysis_mask_r9_7yr_v4.fits’
wmap_temp 0%[ ] 0 --.-KB/s
wmap_tempe 6%[> ] 1.60M 7.88MB/s
wmap_temper 15%[==> ] 3.71M 9.15MB/s
wmap_tempera 24%[===> ] 5.81M 9.50MB/s
wmap_temperat 32%[=====> ] 7.92M 9.72MB/s
wmap_temperatu 41%[=======> ] 10.02M 9.85MB/s
- wmap_temperatur 50%[=========> ] 12.13M 9.94MB/s
</pre>
- wmap_temperatur 50%[=========> ] 12.13M 9.94MB/s
end{sphinxVerbatim}
wmap_temperatur 50%[=========> ] 12.13M 9.94MB/s
- more-to-come:
- wmap_temperature 59%[==========> ] 14.23M 10.0MB/s
</pre>
- wmap_temperature 59%[==========> ] 14.23M 10.0MB/s
end{sphinxVerbatim}
wmap_temperature 59%[==========> ] 14.23M 10.0MB/s
- more-to-come:
- wmap_temperature_ 68%[============> ] 16.33M 10.1MB/s
</pre>
- wmap_temperature_ 68%[============> ] 16.33M 10.1MB/s
end{sphinxVerbatim}
wmap_temperature_ 68%[============> ] 16.33M 10.1MB/s
- more-to-come:
- wmap_temperature_a 76%[==============> ] 18.38M 10.0MB/s
</pre>
- wmap_temperature_a 76%[==============> ] 18.38M 10.0MB/s
end{sphinxVerbatim}
wmap_temperature_a 76%[==============> ] 18.38M 10.0MB/s
- wmap_temperature_an 85%[================> ] 20.48M 10.1MB/s
</pre>
- wmap_temperature_an 85%[================> ] 20.48M 10.1MB/s
end{sphinxVerbatim}
wmap_temperature_an 85%[================> ] 20.48M 10.1MB/s
- map_temperature_ana 94%[=================> ] 22.59M 10.1MB/s
</pre>
- map_temperature_ana 94%[=================> ] 22.59M 10.1MB/s
end{sphinxVerbatim}
map_temperature_ana 94%[=================> ] 22.59M 10.1MB/s
wmap_temperature_an 100%[===================>] 24.01M 10.1MB/s in 2.4s
2024-03-19 15:41:47 (10.1 MB/s) - ‘wmap_temperature_analysis_mask_r9_7yr_v4.fits’ saved [25174080/25174080]
</pre>
wmap_temperature_an 100%[===================>] 24.01M 10.1MB/s in 2.4s
2024-03-19 15:41:47 (10.1 MB/s) - ‘wmap_temperature_analysis_mask_r9_7yr_v4.fits’ saved [25174080/25174080]
end{sphinxVerbatim}
wmap_temperature_an 100%[===================>] 24.01M 10.1MB/s in 2.4s
2024-03-19 15:41:47 (10.1 MB/s) - ‘wmap_temperature_analysis_mask_r9_7yr_v4.fits’ saved [25174080/25174080]
[14]:
wmap_map_I = hp.read_map("wmap_band_iqumap_r9_7yr_W_v4.fits")
By default, input maps are converted to RING ordering, if they are in NESTED ordering. You can otherwise specify nest=True
to retrieve a map is NESTED ordering, or nest=None
to keep the ordering unchanged.
By default, read_map
loads the first column, for reading other columns you can specify the field
keyword.
write_map
writes a map to disk in FITS format, if the input map is a list of 3 maps, they are written to a single file as I,Q,U polarization components:
[15]:
hp.write_map("my_map.fits", wmap_map_I, overwrite=True)
setting the output map dtype to [dtype('>f4')]
Visualization¶
As shown above, mollweide projection with mollview
is the most common visualization tool for HEALPIX maps. It also supports coordinate transformation, coord
does Galactic to ecliptic coordinate transformation, norm='hist'
sets a histogram equalized color scale and xsize
increases the size of the image. graticule
adds meridians and parallels.
[16]:
hp.mollview(
wmap_map_I,
coord=["G", "E"],
title="Histogram equalized Ecliptic",
unit="mK",
norm="hist",
min=-1,
max=1,
)
hp.graticule()
gnomview
instead provides gnomonic projection around a position specified by rot
, for example you can plot a projection of the galactic center, xsize
and ysize
change the dimension of the sky patch.
[17]:
hp.gnomview(wmap_map_I, rot=[0, 0.3], title="GnomView", unit="mK", format="%.2g")
mollzoom
is a powerful tool for interactive inspection of a map, it provides a mollweide projection where you can click to set the center of the adjacent gnomview panel. ## Masked map, partial maps
By convention, HEALPIX uses \(-1.6375 * 10^{30}\) to mark invalid or unseen pixels. This is stored in healpy as the constant UNSEEN
.
All healpy
functions automatically deal with maps with UNSEEN
pixels, for example mollview
marks in grey those sections of a map.
There is an alternative way of dealing with UNSEEN pixel based on the numpyMaskedArray
class, hp.ma
loads a map as a masked array, by convention the mask is 0 where the data are masked, while numpy defines data masked when the mask is True, so it is necessary to flip the mask.
[18]:
mask = hp.read_map("wmap_temperature_analysis_mask_r9_7yr_v4.fits").astype(np.bool_)
wmap_map_I_masked = hp.ma(wmap_map_I)
wmap_map_I_masked.mask = np.logical_not(mask)
Filling a masked array fills in the UNSEEN
value and return a standard array that can be used by mollview
. compressed()
instead removes all the masked pixels and returns a standard array that can be used for examples by the matplotlib hist()
function:
[19]:
hp.mollview(wmap_map_I_masked.filled())
[20]:
plt.hist(wmap_map_I_masked.compressed(), bins=1000);
Spherical Harmonics transforms¶
healpy
provides bindings to the C++ HEALPIX library for performing spherical harmonic transforms. hp.anafast
computes the angular power spectrum of a map:
[21]:
LMAX = 1024
cl = hp.anafast(wmap_map_I_masked.filled(), lmax=LMAX)
ell = np.arange(len(cl))
therefore we can plot a normalized CMB spectrum and write it to disk:
[22]:
plt.figure(figsize=(10, 5))
plt.plot(ell, ell * (ell + 1) * cl)
plt.xlabel("$\ell$")
plt.ylabel("$\ell(\ell+1)C_{\ell}$")
plt.grid()
hp.write_cl("cl.fits", cl, overwrite=True)
Gaussian beam map smoothing is provided by hp.smoothing
:
[23]:
wmap_map_I_smoothed = hp.smoothing(wmap_map_I, fwhm=np.radians(1.))
hp.mollview(wmap_map_I_smoothed, min=-1, max=1, title="Map smoothed 1 deg")
For more information see the HEALPix primer