From b638199372b9c5016c2e0eb1da6e07b96eae2be3 Mon Sep 17 00:00:00 2001
From: RichardScottOZ <72196131+RichardScottOZ@users.noreply.github.com>
Date: Fri, 30 Oct 2020 15:50:47 +1030
Subject: [PATCH] Update conjugate_gradient2.ipynb

Couple of typos.
---
 1804_Conjugate_gradient_inversion/conjugate_gradient2.ipynb | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/1804_Conjugate_gradient_inversion/conjugate_gradient2.ipynb b/1804_Conjugate_gradient_inversion/conjugate_gradient2.ipynb
index f585381..8629f08 100644
--- a/1804_Conjugate_gradient_inversion/conjugate_gradient2.ipynb
+++ b/1804_Conjugate_gradient_inversion/conjugate_gradient2.ipynb
@@ -8,7 +8,7 @@
     "\n",
     "An exploration of the conjugate gradient solver. Considering the problem of getting reflectivity from a seismic trace, given the wavelet.\n",
     "\n",
-    "This was the topic of a previous tutorial which used the numpy and scipy matrix solvers. This notebook solves same problem the conjugate gradient method and compares the result to the matrix solver results.\n",
+    "This was the topic of a previous tutorial which used the numpy and scipy matrix solvers. This notebook solves same problem by the conjugate gradient method and compares the result to the matrix solver results.\n",
     "\n",
     "We'll be working on a model, so we know the correct answer from the start."
    ]
@@ -685,7 +685,7 @@
    "source": [
     "## Estimate the model using the preconditioned conjugate gradient method\n",
     "\n",
-    "I will use the previous unconditioned results and attempt to concentrate amplitude in the locations that already have large amplitude. The  preconditioner is to scale by sqrt(previous result smoothed amplitude^2). This scalar is just the RMS of a 3 point sliding window. The effect will be to make large amplitudes larger and small ampltudes smaller in the estimated model.\n",
+    "I will use the previous unconditioned results and attempt to concentrate amplitude in the locations that already have large amplitude. The  preconditioner is to scale by sqrt(previous result smoothed amplitude^2). This scalar is just the RMS of a 3 point sliding window. The effect will be to make large amplitudes larger and small amplitudes smaller in the estimated model.\n",
     "\n",
     "Compute R the right preconditioning diagonal matrix.\n",
     "\n",
@@ -900,7 +900,7 @@
    "source": [
     "def mv(v):\n",
     "    \"\"\"\n",
-    "    The fordward operator, convolution.\n",
+    "    The forward operator, convolution.\n",
     "    \"\"\"\n",
     "    return np.convolve(v, wavelet, mode='same')\n",
     "\n",