diff --git a/CMB_ACT_Likelihood.ipynb b/CMB_ACT_Likelihood.ipynb index 0961d79..f20fd5e 100644 --- a/CMB_ACT_Likelihood.ipynb +++ b/CMB_ACT_Likelihood.ipynb @@ -722,7 +722,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Repeat for deep patch (wherever you see data being accessed from the likelihood objects in the code above, change the number \"130\" to a \"0\"). Comment on any similarities/differences.\n", + "EXERCISE: Repeat for deep patch (wherever you see data being accessed from the likelihood objects in the code above, change the number \"130\" to a \"0\"). Comment on any similarities/differences.\n", "\n", "Optional next step: use the tools later in the notebook to see which cosmological parameters are better-constrained by which patch" ] @@ -1030,7 +1030,7 @@ "metadata": {}, "source": [ "### We don't actually want to read in the data every time. \n", - "EXCERCISE: Change the likelihood function to only read in the data the first time it is called." + "EXERCISE: Change the likelihood function to only read in the data the first time it is called." ] }, { @@ -1054,7 +1054,7 @@ "metadata": {}, "source": [ "### This code is really ugly (and slow)!\n", - "EXCERCISE: Write functions/modules to speed up the MCMC code above." + "EXERCISE: Write functions/modules to speed up the MCMC code above." ] }, { @@ -1349,14 +1349,14 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Look at correlation between other parameters." + "EXERCISE: Look at correlation between other parameters." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Compare these 1D and 2D with the equivalent for the chain you ran before." + "EXERCISE: Compare these 1D and 2D with the equivalent for the chain you ran before." ] }, { @@ -1387,7 +1387,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Modify your code above to compute the acceptance/rejection ratio while the steps are being taken." + "EXERCISE: Modify your code above to compute the acceptance/rejection ratio while the steps are being taken." ] }, { @@ -1417,7 +1417,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Modify your stepping function to take a covariance matrix (determined from a shorter run of the chain) and to step using this covariance matrix." + "EXERCISE: Modify your stepping function to take a covariance matrix (determined from a shorter run of the chain) and to step using this covariance matrix." ] }, { diff --git a/CMB_School_Part_02.ipynb b/CMB_School_Part_02.ipynb index aa07d76..7e620fc 100644 --- a/CMB_School_Part_02.ipynb +++ b/CMB_School_Part_02.ipynb @@ -155,7 +155,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Plot a histogram of the AGN brightness values and a histogram of the DSFG brightness values. How do these two distributions differ?" + "EXERCISE: Plot a histogram of the AGN brightness values and a histogram of the DSFG brightness values. How do these two distributions differ?" ] }, { @@ -277,7 +277,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: In reality the cluster radii vary from cluster to cluster. For reference the radius depends on redshift and mass. The number of clusters as a function of mass and redshift is called the cluster 'mass function' which is a sensitive cosmological proble. \n", + "EXERCISE: In reality the cluster radii vary from cluster to cluster. For reference the radius depends on redshift and mass. The number of clusters as a function of mass and redshift is called the cluster 'mass function' which is a sensitive cosmological proble. \n", "\n", "To enhance the realism of our sims, divide the simulated cluster sample into an extremely large radius sample (1 cluster with a 30 arcminute radius, comparable to the Coma cluster; the largest cluster on the sky), a large radius bin (10%) clusters with 5 arcminute radius), a medium bin (30%) with 2 arcminute radius, and a small bin (60% with 0.5 arcminute radius). " ] diff --git a/CMB_School_Part_03.ipynb b/CMB_School_Part_03.ipynb index 2ee668c..9538389 100644 --- a/CMB_School_Part_03.ipynb +++ b/CMB_School_Part_03.ipynb @@ -163,7 +163,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: There are three typical beam sizes for CMB experiments: 1) large angular scale B-mode observatories which have ~30 arcminute beams; 2) medium scale observatories such as the Planck satellite that has a ~3 arcminute beam, and 3) high resolution observatories such as ACT and SPT that have ~1 arcminute beams. Convolve your map with each of these cases and compare the differences." + "EXERCISE: There are three typical beam sizes for CMB experiments: 1) large angular scale B-mode observatories which have ~30 arcminute beams; 2) medium scale observatories such as the Planck satellite that has a ~3 arcminute beam, and 3) high resolution observatories such as ACT and SPT that have ~1 arcminute beams. Convolve your map with each of these cases and compare the differences." ] }, { @@ -254,7 +254,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: The above simulation has both atmospheric and detector $1/f$ noise. To understand these components seperately, turn each off and make plots of each component seperately. Comment on how these look. \n", + "EXERCISE: The above simulation has both atmospheric and detector $1/f$ noise. To understand these components seperately, turn each off and make plots of each component seperately. Comment on how these look. \n", "\n", "Optional Part 2: Plot the absolute value of the 2-d FFT of these maps and note how the non-white component of the noise is localized. \n", "\n", @@ -324,7 +324,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Find the minimum $N_{mask}$ that effectively whitens the noise spectrum with the bad $1/f$ turned on. See how this filter works for atmospheric noise. Comment on whether this filter works and whether it is a good choice for atmospheric noise." + "EXERCISE: Find the minimum $N_{mask}$ that effectively whitens the noise spectrum with the bad $1/f$ turned on. See how this filter works for atmospheric noise. Comment on whether this filter works and whether it is a good choice for atmospheric noise." ] }, { diff --git a/CMB_School_Part_04.ipynb b/CMB_School_Part_04.ipynb index 0a278fc..20113fd 100644 --- a/CMB_School_Part_04.ipynb +++ b/CMB_School_Part_04.ipynb @@ -176,7 +176,7 @@ "metadata": {}, "source": [ "This shows our simulated map with a cosine window applied to eliminate edge effects. It is obvious from this map that we are suppressing the signal here.\n", - "EXCERCISE: There are an huge number of well studied windows with various combinations of properties. Some minimize mode coupling, others minimize signal loss, while others maximize some combination of the two. Find the wikipedia article on Fourier transform windows, choose one of your favorites and implement it as an option. Compare the impact of the new window compared to the simple cosine window on the map." + "EXERCISE: There are an huge number of well studied windows with various combinations of properties. Some minimize mode coupling, others minimize signal loss, while others maximize some combination of the two. Find the wikipedia article on Fourier transform windows, choose one of your favorites and implement it as an option. Compare the impact of the new window compared to the simple cosine window on the map." ] }, { @@ -298,7 +298,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Comment on how the measured power sepctrum (blue) compares to the input spectrum (green). Why are they different? What features are common to both? How does your alternative window funciton affect the measured spectrum compared to the default?" + "EXERCISE: Comment on how the measured power sepctrum (blue) compares to the input spectrum (green). Why are they different? What features are common to both? How does your alternative window funciton affect the measured spectrum compared to the default?" ] }, { @@ -442,7 +442,7 @@ "metadata": {}, "source": [ "This plot shows the estimate of the CMB power spectrum after correcting for the multiplicative bias (transfer function) in yellow. In addition we show (red) the input CMB power spectrum, (green) the average of the signal only simulations, (blue, lower) the transfer function, and (blue, upper) the naive power spectrum of our map. Consider how all these curves relate to creating the yellow estimate.\n", - "EXCERCISE: Why does the green curve look so much like the red curved but surpressed? What is the meaning of the lower (and smoother) blue curve?" + "EXERCISE: Why does the green curve look so much like the red curved but surpressed? What is the meaning of the lower (and smoother) blue curve?" ] }, { @@ -545,7 +545,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: What is the yellow curve? How is it generated, how is it used to get the green curve?" + "EXERCISE: What is the yellow curve? How is it generated, how is it used to get the green curve?" ] }, { @@ -669,7 +669,7 @@ "metadata": {}, "source": [ "And there you have it. That is how you compute a CMB power spectrum and error bars. If you want to fit cosmology to these data you can re-run CAMB varying cosmological parameters and compute the likelihood difference between these models and the data. This is left as an exercise.\n", - "EXCERCISE: Why dosn't the red curve (input CMB spectrum) fall wihin the error bars of our simulated data set above $\\ell \\sim 2500$?" + "EXERCISE: Why dosn't the red curve (input CMB spectrum) fall wihin the error bars of our simulated data set above $\\ell \\sim 2500$?" ] }, { @@ -683,7 +683,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: One can worry that the transfer funciton calcuation depends on the input spectrum used in its calculation. Modify the spectrum used in calculating the transfer funciton to better match the spectrum in the simualtion by adding a constant in quadrature to the input spectrum. Comment on weather this changes the result or not." + "EXERCISE: One can worry that the transfer funciton calcuation depends on the input spectrum used in its calculation. Modify the spectrum used in calculating the transfer funciton to better match the spectrum in the simualtion by adding a constant in quadrature to the input spectrum. Comment on weather this changes the result or not." ] }, { diff --git a/CMB_School_Part_06.ipynb b/CMB_School_Part_06.ipynb index 48277ef..9ff205d 100644 --- a/CMB_School_Part_06.ipynb +++ b/CMB_School_Part_06.ipynb @@ -223,7 +223,7 @@ "source": [ "Notice that the power spectrum is biased high on large scales - as we expected it to be.\n", "How might you remove that large scale power before computing the power spectrum?\n", - "EXCERCISE: Write code to filter out the large scale modes ell < ell_min before computing the power spectrum and show the spectra for a few different values for ell_min. Also, can you think about how to remove the ringing the power on small scales? Discuss the shape of the apodising window and the large scale power leaking to small scales." + "EXERCISE: Write code to filter out the large scale modes ell < ell_min before computing the power spectrum and show the spectra for a few different values for ell_min. Also, can you think about how to remove the ringing the power on small scales? Discuss the shape of the apodising window and the large scale power leaking to small scales." ] }, { @@ -239,7 +239,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: We cut out a particular patch from the ACT strip. Cut out different patches and view them. By using the power spectrum code in the modules file, compute both the auto and the cross spectra for this map. Discuss how you would estimate the error bars on this power spectrum, by comparing the auto with the cross power spectra." + "EXERCISE: We cut out a particular patch from the ACT strip. Cut out different patches and view them. By using the power spectrum code in the modules file, compute both the auto and the cross spectra for this map. Discuss how you would estimate the error bars on this power spectrum, by comparing the auto with the cross power spectra." ] }, { @@ -255,7 +255,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Given the value of the PS shown here on small scales, we can estimate the noise level in the ACT maps for this season. Plot noise curves for a few values and show here, and do a simple \"chi-by-eye\" fit for the noise level." + "EXERCISE: Given the value of the PS shown here on small scales, we can estimate the noise level in the ACT maps for this season. Plot noise curves for a few values and show here, and do a simple \"chi-by-eye\" fit for the noise level." ] }, { @@ -278,7 +278,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Plot the 2D power spectrum, both of your input theory+ noise and the ACT data. What does that tell you about real-world noise from CMB Experiments?" + "EXERCISE: Plot the 2D power spectrum, both of your input theory+ noise and the ACT data. What does that tell you about real-world noise from CMB Experiments?" ] }, { @@ -353,7 +353,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: The spectrum is biased high from point sources on all scales. Use the code from the previous notebook to determine the noise bias for this spectrum and remove it." + "EXERCISE: The spectrum is biased high from point sources on all scales. Use the code from the previous notebook to determine the noise bias for this spectrum and remove it." ] }, { @@ -374,7 +374,7 @@ } }, "source": [ - "EXCERCISE: Apply the techniques from Part Five and search for point sources and SZ clusters in the map" + "EXERCISE: Apply the techniques from Part Five and search for point sources and SZ clusters in the map" ] }, { diff --git a/CMB_School_Part_07.ipynb b/CMB_School_Part_07.ipynb index 1124a65..5f69278 100644 --- a/CMB_School_Part_07.ipynb +++ b/CMB_School_Part_07.ipynb @@ -411,7 +411,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Add point sources and the SZ to these maps, next convolve with a beam, and add instrumental noise. Assume the point sources have a fixed $3\\%$ polarization fraction with random polarization angles and the SZ is fully unpolarized. The noise in the polarization maps is $\\sqrt 2$ larger than the temperature noise." + "EXERCISE: Add point sources and the SZ to these maps, next convolve with a beam, and add instrumental noise. Assume the point sources have a fixed $3\\%$ polarization fraction with random polarization angles and the SZ is fully unpolarized. The noise in the polarization maps is $\\sqrt 2$ larger than the temperature noise." ] }, { @@ -513,7 +513,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Generate a polarization field dominated by B-mode polarization and plot the polarization vectors on top of the B-mode map. What is the difference between the E-mode field and the B-mode field?" + "EXERCISE: Generate a polarization field dominated by B-mode polarization and plot the polarization vectors on top of the B-mode map. What is the difference between the E-mode field and the B-mode field?" ] }, { @@ -529,7 +529,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Calculate the BB power spectrum using the code and Monte Carlo techniques we explored in Part Four. Compare your results to the input spectra. Make sure to include a window (i.e. pixel mask). \n", + "EXERCISE: Calculate the BB power spectrum using the code and Monte Carlo techniques we explored in Part Four. Compare your results to the input spectra. Make sure to include a window (i.e. pixel mask). \n", "\n", "\n", "hint: You can make use of the `QU2EB()` function below to convert Q and U maps to E and B maps. Because real data would be provided as Q and U maps, a pixel mask should be applied to the Q and U maps before converting to E and B." @@ -646,7 +646,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Create a simulated map with $C_\\ell^{BB} = 0$ Calculate the BB power spectrum of this map using the estimates for the multiplicitive and addative bias you calculated in the previous excercise with non-zero $C_\\ell^{BB}$. Why isn't your result consistent with $C_\\ell^{BB} = 0$ " + "EXERCISE: Create a simulated map with $C_\\ell^{BB} = 0$ Calculate the BB power spectrum of this map using the estimates for the multiplicitive and addative bias you calculated in the previous excercise with non-zero $C_\\ell^{BB}$. Why isn't your result consistent with $C_\\ell^{BB} = 0$ " ] }, { @@ -669,7 +669,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Calculate the TE power spectrum by adapting the techniques from Part Four." + "EXERCISE: Calculate the TE power spectrum by adapting the techniques from Part Four." ] }, { @@ -827,7 +827,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Calculate the TT, EE, BB, TE, TB, and EB power spectra using the code and techniques we explored in Part Four but with the Kendrick-method for B-modes. Compare your results to the input spectra and to the spectra you produced with the naive E-B separation. Pay special attention to the angular scale dependence of the differences." + "EXERCISE: Calculate the TT, EE, BB, TE, TB, and EB power spectra using the code and techniques we explored in Part Four but with the Kendrick-method for B-modes. Compare your results to the input spectra and to the spectra you produced with the naive E-B separation. Pay special attention to the angular scale dependence of the differences." ] }, { @@ -859,7 +859,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "ADVANCED EXCERCISE: Take the `lens_map` function from the lensing notebook (part 10) and lens the Q and U maps of a polarization field that has no B-mode contribution. Plot the resulting B-mode power spectrum and compare to the CAMB B-mode power spectrum." + "ADVANCED EXERCISE: Take the `lens_map` function from the lensing notebook (part 10) and lens the Q and U maps of a polarization field that has no B-mode contribution. Plot the resulting B-mode power spectrum and compare to the CAMB B-mode power spectrum." ] }, { diff --git a/CMB_School_Part_08.ipynb b/CMB_School_Part_08.ipynb index b7cae59..0011357 100644 --- a/CMB_School_Part_08.ipynb +++ b/CMB_School_Part_08.ipynb @@ -176,7 +176,7 @@ "metadata": {}, "source": [ "The first thing we notice is that we shouldn't just be taking the model spectrum at that bin, but we should be binning the theory.\n", - "EXCERCISE: Write a function to bin the theory over the same ell range as the binned data." + "EXERCISE: Write a function to bin the theory over the same ell range as the binned data." ] }, { @@ -369,7 +369,7 @@ "metadata": {}, "source": [ "### We don't actually want to read in the data every time. \n", - "EXCERCISE: Change the likelihood function to only read in the data the first time it is called." + "EXERCISE: Change the likelihood function to only read in the data the first time it is called." ] }, { @@ -411,7 +411,7 @@ "metadata": {}, "source": [ "### This code is really ugly (and slow)!\n", - "EXCERCISE: Write functions/modules to speed up the MCMC code above." + "EXERCISE: Write functions/modules to speed up the MCMC code above." ] }, { @@ -443,7 +443,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Modify your code above to compute the acceptance/rejection ratio while the steps are being taken." + "EXERCISE: Modify your code above to compute the acceptance/rejection ratio while the steps are being taken." ] }, { @@ -479,7 +479,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Modify your stepping function to take a covariance matrix (determined from a shorter run of the chain) and to step using this covariance matrix." + "EXERCISE: Modify your stepping function to take a covariance matrix (determined from a shorter run of the chain) and to step using this covariance matrix." ] }, { diff --git a/CMB_School_Part_09.ipynb b/CMB_School_Part_09.ipynb index e098612..67120b3 100644 --- a/CMB_School_Part_09.ipynb +++ b/CMB_School_Part_09.ipynb @@ -218,7 +218,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Make a zoomed in plot of the scan strategies above by adjusting the x limit for the plot to be much smaller. Discribe the scan strategy based on what you see in these plots." + "EXERCISE: Make a zoomed in plot of the scan strategies above by adjusting the x limit for the plot to be much smaller. Discribe the scan strategy based on what you see in these plots." ] }, { @@ -284,7 +284,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Explain the spikes in the TOD and discuss weather the RMS of the TOD makes sense." + "EXERCISE: Explain the spikes in the TOD and discuss weather the RMS of the TOD makes sense." ] }, { @@ -298,7 +298,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "OPTIONAL EXCERCISE: Add in the impact of detector time constants to smear out these TODs. This excercise is best done aftern completig the rest of the notebook so its impact can be propigatd through this entire code to maps. This will allow you to model the impact of this sytematic on effect.\n", + "OPTIONAL EXERCISE: Add in the impact of detector time constants to smear out these TODs. This excercise is best done aftern completig the rest of the notebook so its impact can be propigatd through this entire code to maps. This will allow you to model the impact of this sytematic on effect.\n", "\n", "NOTE: [this wikipedia article](https://en.wikipedia.org/wiki/Exponential_smoothing) is useful in modeling exponetial filters (e.g., detector time cosntants) This apporach can be implemented with arrays using the np.roll funciton." ] @@ -409,7 +409,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Typical detectors have noises of $\\sim 300 \\mu$K-$\\sqrt{s}$. Make a plot of what a time stream looks like with this instantaneous sensitivity so you can get a feel for what real data looks like." + "EXERCISE: Typical detectors have noises of $\\sim 300 \\mu$K-$\\sqrt{s}$. Make a plot of what a time stream looks like with this instantaneous sensitivity so you can get a feel for what real data looks like." ] }, { @@ -580,7 +580,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Do you see comon structures in this map and the input sky map?" + "EXERCISE: Do you see comon structures in this map and the input sky map?" ] }, { @@ -594,7 +594,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Remake the TOD and this map setting the 1/f knee to be absurly small. Compare the new result to the originaly input map?" + "EXERCISE: Remake the TOD and this map setting the 1/f knee to be absurly small. Compare the new result to the originaly input map?" ] }, { @@ -617,7 +617,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: reset the 1/f knee to its original value. Now filter the time-series using a high-pass filter with a knee frequency set to whiten the TODs. Remake the map and compare the new result to the input map. What is different?" + "EXERCISE: reset the 1/f knee to its original value. Now filter the time-series using a high-pass filter with a knee frequency set to whiten the TODs. Remake the map and compare the new result to the input map. What is different?" ] }, { @@ -927,14 +927,14 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Compare the cross-linked and non-crosslinked maps. The horizontal strips in the non-crosslinked maps are refered to (uncreatively) as stripes." + "EXERCISE: Compare the cross-linked and non-crosslinked maps. The horizontal strips in the non-crosslinked maps are refered to (uncreatively) as stripes." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Make plots of these maps in 2-d fourier space" + "EXERCISE: Make plots of these maps in 2-d fourier space" ] }, { @@ -957,7 +957,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Filter the non-crosslinked map to remove the stripes. The can be accomplised by masking the relevent modes in your 2d fourier plot before inverse fouerier transforming to create the map." + "EXERCISE: Filter the non-crosslinked map to remove the stripes. The can be accomplised by masking the relevent modes in your 2d fourier plot before inverse fouerier transforming to create the map." ] }, { diff --git a/CMB_School_Part_10.ipynb b/CMB_School_Part_10.ipynb index 00e6354..84a40fa 100644 --- a/CMB_School_Part_10.ipynb +++ b/CMB_School_Part_10.ipynb @@ -398,7 +398,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "EXCERCISE: Compare the power spectrum of the lensed CMB map (no beam, no noise) with the unlensed CMB map (no beam, no noise). Can you clearly see what features lensing induces? If it's too noisy, you could try Monte Carloing multiple realizations and averaging over them?" + "EXERCISE: Compare the power spectrum of the lensed CMB map (no beam, no noise) with the unlensed CMB map (no beam, no noise). Can you clearly see what features lensing induces? If it's too noisy, you could try Monte Carloing multiple realizations and averaging over them?" ] }, {